Peter Brooke

Snapshot and testable edge thesis

Core philosophy

Brooke's direct voice begins with humility about agency and uncertainty. He said the entrepreneur is the protagonist, the investor an enabler; called success random; denied being a risk-taker; and described reading people as the job. [investor-stated · strong] Those statements support selection, assistance, and portfolio humility—not passivity or proof of superior judgment. [researcher inference · strong] CHM transcript, printed pp. 6, 11–12, and 22

His most specific selection chain was to seek a growing sector, defensible technology, management able to commercialize it, current outside technical expertise, and dissent strong enough to challenge the financial sponsor. [investor-stated · strong] The public record does not establish how consistently the chain was applied or whether it outperformed. [researcher inference · strong] CHM transcript, printed pp. 36–40

[researcher inference · moderate] The operating system is therefore conditional: source where knowledge is real; use an instrument the company can survive; choose candid, adaptable people; keep negative information moving; intervene only through an explicit governance right and measurable mechanism; separate mission from economics; and change the vehicle when capital, geography, or incentives no longer fit. That synthesis is reconstructed from the record, not a Brooke-authored checklist. Written-corpus decision playbook · evidence-backed deal rules

Sourcing

• Relationship before transaction. [investor-stated · moderate] Brooke described using the bank's broad borrower funnel to observe companies before retaining the strongest relationships, and the memoir's visible passages reuse contacts formed at the bank, Bessemer, and Tucker Anthony. No source-to-close or source-to-return denominator survives. Concord transcript, policy-paper paragraph · A Vision, printed pp. 9, 25, and 27
• Map comparative advantage. [investor-stated · strong] Brooke's regional and international rule was to finance where technical, market, or institutional knowledge created an advantage. The rule narrows the search space; it does not prove that a cluster, university, or famous sponsor creates alpha. Concord transcript, paragraph beginning “It seemed to me”
• Use a locally accountable source. [investor-stated · moderate] The global model favored a “local champion” with market knowledge and local relationships. Japan and Tang's Hong Kong account show that a champion without aligned economics, authority, and committed support is not a repeatable sourcing engine. A Vision, printed pp. 65–71 and 97–98 · Tang transcript, printed pp. 69–72
• Do not back-attribute the later TA machine. [retrospective witness · strong] Jacqui Morby says she built the magazine/Dun & Bradstreet/cold-call workflow, took leads to Kevin Landry, and that Brooke was not materially involved in it. Brooke's supported contribution was hiring and cultural permission. Morby exact MP3, 15:15–17:28 and 32:59–35:18

Decision use — [researcher inference · moderate]: name the opportunity's knowledge advantage, source owner, relationship history, information acquired before the financing, conflicts, and conversion denominator. Kill the “proprietary” claim when the advantage is only reputation or geography. Sourcing model and boundary cases

Picking

Brooke's best-supported founder screen is behavioral rather than biographical. He said he tested business-plan assumptions through founders' responses: candor about an unknown improved trust; evasion damaged the future working relationship. [investor-stated · strong] CHM transcript, printed pp. 22–24

Red flags — [researcher inference · moderate]: evasive founders; no independent validation of the technical claim; cash needs that exceed the proposed instrument and reserves; founder succession without an executable plan; a local manager without authority; and an “operating value” story without owner, cost, baseline, or milestone. Wang, Unitrode, Aspen, Japan, and INF labs

Deal and portfolio mechanics

• Instrument. [investor-stated · strong] In 1975 Brooke described notes plus common equity as an alternative to straight equity when exits were weak; Pete Bancroft immediately supplied the start-up cash-flow boundary. 1975 article, printed/PDF p. 6
• Ownership and price. [researcher inference · insufficient public record] No universal Brooke ownership target, valuation rule, liquidation preference, dilution policy, or price discipline surfaced. [contemporaneous record · strong] A Unitrode proxy records less than 1% beneficial ownership in 1995, including options, but not 1961 cost basis or holding continuity. Unitrode 1995 proxy, “Security Ownership of Certain Beneficial Owners and Management” table
• Reserves and follow-ons. [researcher inference · insufficient public record] Federal Express supplies a vivid last-round conviction story, but Brooke credits Charlie Lea and another TA colleague; the record lacks the partnership, check, reserve, concentration, and proceeds. Lea oral history, Tape 4 00:02:36–00:04:06 · CHM, printed pp. 38–40
• Portfolio construction. [investor-stated · moderate] Diversify stage exposure when the capital provider requires prudence, and separate early venture, expansion, corporate accounts, country funds, and buyouts. The INF postmortem is evidence that an undisciplined early-stage overlay can corrupt a global mandate. CHM, printed pp. 6 and 22–24 · A Vision, printed pp. 114–117
• Exits. [investor-stated · strong] Brooke's 1970s realization drought separates company progress from realizable return; a reopened IPO market in 1979–1981 was part of the outcome mechanism. Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s”

[researcher inference · strong] The public record supports an instrument-and-mandate philosophy more strongly than a complete ownership, reserve, concentration, or sell discipline. Any modern use should begin with the missing fields rather than infer them from famous outcomes. Deal-forensics boundaries

Vehicle and incentive fit

Brooke's memoir criticizes manager compensation that can enrich the manager without commensurate fund success and describes retained corporate economics and shared ownership as institution-building tools. [investor-stated · moderate] A Vision, printed pp. 54, 79, 166, and 187–188 Without agreements, fee offsets, carry allocations, clawbacks, GP commitments, and LP cash flows, alignment remains a stated design rather than verified effect. [researcher inference · strong] Vehicle-economics retrieval ledger

Team attribution and work with founders

[researcher inference · strong] Brooke's most supportable founder contribution is bounded enablement: relationship access, financing architecture, recruitment, strategy, and institutional support. It is not sole causation, and Wang shows that long service cannot substitute for control or adaptation. People and decision map · Wang decision lab

Views on the craft

• Power law and luck. Brooke called success random and favored diversification. [investor-stated · strong] CHM, printed pp. 11 and 22–24 The public deal record contains only one unranked numerical deal-return estimate—[investor-stated · provisional] reported: ~50x (Boston Globe obituary, 2020)—and no Brooke-personal denominator. [researcher inference · strong] Economic sensitivity
• Access versus selection. [researcher inference · moderate] Relationships and clusters created access; observation, candor, and technical dissent were intended to select. Missing funnel data prevent measuring either contribution. Concord transcript, paragraphs beginning “The department that I started” and “It seemed to me” · reconciliation
• Fund size and mandate. [investor-stated · moderate] Separate stages and vehicles rather than let a global mission create uncontrolled early-stage exposure. INF is the negative control. A Vision, printed pp. 114–117
• Governance. [investor-stated · strong] Investors are useful in strategy, management structure, recruiting, financing, and sale negotiation, but cannot manufacture a good company from a bad idea or team. CHM, printed pp. 30–35
• Operating value and exits. Durable operating improvement, not leverage alone, is Brooke's claimed engine. [investor-stated · strong] WBUR 2007, Brooke 10:00–11:16 Exit markets remain an independent return input. [researcher inference · strong] Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s”
• Great investor. [researcher inference · moderate] The record implies a selector and institution-builder who subordinates ego, develops people, adapts vehicles, and admits failure. It does not support a timeless seer, sole dealmaker, or audited personal performance claim. Profile edge test · evidence boundary

Signature bets and decision process

The decision evidence belongs in the no-hindsight labs rather than in retrospective victory stories. [researcher inference · strong] Decision-lab method

1. Bank technology lending — can credit structure buy observation without hiding default risk?
2. Wang — when do relationship access and board service fail to create control?
3. Unitrode — what evidence justifies management change, and who actually owns it?
4. Federal Express — what new evidence justifies reserve concentration, and whose decision was it?
5. Biogen — can the investment be evaluated without the later multiple?
6. Aspen Technology — can each claimed value-creation mechanism be pre-registered and measured?
7. Japan — are local/central rights and stop gates investable before launch?
8. International Network Fund — can a global pool absorb different stage and manager risks?
9. Hong Kong concept — are sponsor support and co-investment documented obligations or hopes?

[researcher inference · strong] The cases show a decision system more clearly than they show personal economics. The correct replay order is ex-ante information → decision rights and instrument → milestone/follow-on → actual outcome → counterfactual, never outcome first. Nine no-hindsight labs · economic boundary

Mental models and maxims

[researcher inference · strong] The reconciled inventory contains thirteen supported or explicitly bounded models; synonyms are merged, credit-as-information remains distinct from instrument fit, and mission/return scorecards remain cross-cutting doctrine rather than a model because three Brooke-attributable applications did not survive. Model reconciliation and exhausted routes

1. Use credit as an information option only when observation changes the next decision.
2. Match the instrument to cash stage; separate security protection from company risk.
3. Map comparative advantage, then build the relationship funnel around it.
4. Test candor and commercializing management, not founder polish.
5. Pair a local champion with shared standards, explicit authority, and contracted support.
6. Make negative information and outside dissent travel.
7. Enable through the board without confusing influence with control.
8. Pre-register operating value; do not substitute leverage or narrative.
9. Separate mandates and risk budgets; treat exit liquidity as a portfolio input.
10. Develop whole-job investors and make team succession visible.
11. Match vehicle, incentives, authority, and duration to the strategy.
12. Ledger person, vehicle, instrument, and decision separately.
13. Predefine defeat, follow-on gates, and the learning budget.

Evolution and contradictions

Three contradictions are productive analytical tests, not drafting defects. [researcher inference · strong] Dated evidence table

1. Diversification versus conviction — [researcher inference · strong]. Brooke called success random and favored prudent stage diversification, yet praised “faith” and a concentrated final Federal Express follow-on he did not personally lead. CHM, printed pp. 11, 22–24, and 38–40
2. Investor humility versus active intervention — [researcher inference · strong]. He called investors enablers while describing management and strategy work at Wang, Unitrode, and Adams Russell. The coherent boundary is help on a viable company, not omnipotence—but board records are missing. CHM, printed pp. 11–12 and 27–35
3. Local autonomy versus integrated control — [researcher inference · strong]. The network needed local information but moved toward common committee standards; Japan and Tang show that neither decentralization nor sponsorship alone solved authority and support. A Vision, printed pp. 97–113 · Tang, printed pp. 69–72

Failures and limits

Historical operating environment

• 1958 statute — [contemporaneous record · strong]. The Small Business Investment Act created an SBIC framework. 1958 statute Bank method — [investor-stated · moderate]. Brooke later recalled a distinct program using assignable federal contracts when young technical firms lacked conventional assets. Concord transcript, paragraphs beginning “The department that I started” and “The policy paper that I put”
• Defense, electronics, and Route 128 — [investor-stated · moderate]. Military R&D contracts and university-linked technical labor made New England unusually legible to a local lender; later semiconductor migration and minicomputer decline show that a regional advantage can expire. CHM, printed pp. 15–21 · Concord transcript, paragraphs beginning “The leaders of industry” and “That advance obsoleted”
• 1970s illiquidity — [contemporaneous record · strong]. Weak new issues, diminished institutional supply, longer holds, and follow-on needs framed the note-plus-common response. 1975 article, printed/PDF pp. 5–7 [investor-stated · strong] Brooke separately remembered roughly a decade without substantial realizations. Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s”
• Pension regulation — [contemporaneous record · strong]. The 1979 ERISA prudence rule evaluated an investment in portfolio context and diversification; it should not be reduced to “venture became legal.” 44 FR 37221
• Cross-border institution building — [researcher inference · strong]. French public institutions, corporate sponsors, IFC, local managers, currency exposure, thin exits, and uneven legal/governance systems shaped what global venture could mean. Vehicle evidence [documented behavior · strong] A September 1996 IFC retrospective analysis reported: −5.0% aggregate IRR for 21 early funds (IFC Note 90, printed p. 2, 1996; observation as of June 1983). [researcher inference · strong] The report does not identify INF as part of that set. INF boundary

[researcher inference · strong] Brooke’s strategy cannot be lifted out of these conditions. Contract collateral, family-office apprenticeship, thin early fund competition, public-institution sponsorship, and later exit openings were enabling inputs, not incidental scenery or causally identified returns. Historical record · vehicle boundaries

Modern VC translation (as of 2026-08-18)

The 2026 NVCA Yearbook reports that 487 $100-million-plus rounds were 3.2% of 2025 deal count but 67% of value, while two-thirds of unicorn IPOs priced below their last private valuation. [contemporaneous record · strong] 2026 NVCA Yearbook Those observations do not validate a Brooke rule; they make denominator, refinancing, reserve, and exit assumptions more important. [researcher inference · strong] Mandate-and-exit model

[researcher inference · insufficient public record] Mission and financial scorecards remain a cross-cutting anti-confusion rule, not a supported Brooke model: the searched record has aspiration, manager-reported fund economics, and public counterargument, but not three Brooke-attributable applications. Exhausted-route ledger

Comparison with completed peers

[researcher inference · strong] These three comparators are defensible because all are completed pioneer profiles with different capital forms and overlapping problems of technical access, patient ownership, governance, and attribution. The comparison is of mechanisms—not rank or return. Doriot canon · Rockefeller canon · Dennis canon

Comparability boundary — [researcher inference · strong]: Doriot has a much stronger portfolio denominator; Rockefeller and Dennis have better documented personal/family-capital transitions in selected cases. None supplies a directly comparable Brooke-personal return series, so the table does not rank performance. Doriot denominator · Rockefeller capital record · Dennis capital record · Brooke evidence boundary

The strongest case against greatness

[researcher inference · strong] The skeptical thesis is that Brooke was a talented narrator and institutional connector whose reputation combines inherited institutional access, partner work, successor performance, favorable liquidity windows, and selected survivor stories without enough public evidence to demonstrate repeatable personal investing skill. Edge decomposition · searched-denominator boundary

1. Access and privilege. [researcher inference · moderate] Exeter, Harvard/HBS, Bank of Boston, Bessemer/Phipps, Tucker Anthony, government and corporate sponsors supplied unusually rich access and capital. The edge may be access conversion rather than selection. CHM, printed pp. 3–27
2. Team misattribution. [researcher inference · strong] Lea owns the Federal Express lead; Landry and Morby own later TA investing/sourcing; Anagnostopoulos claims material Advent design/capital/recruiting; Brown appears in Aspen and succession records; local managers executed abroad. People map
3. Missing denominator and survivorship. [researcher inference · strong] The public name set is 17/17 searched and 0/17 fully reconstructed; passes, write-offs, reserves, ownership, and cash flows are absent. Denominator
4. One-outlier and successor laundering. The only named numerical deal-return estimate is [investor-stated · provisional] reported: ~50x (Boston Globe obituary, 2020). No public record connects it to Brooke-specific underwriting or cash flow, and later Advent scale or GPE III cannot be assigned backward. [researcher inference · strong] Biogen lab · fund controls
5. Era and exit effects. Brooke reports a decade-long realization drought followed by an IPO reopening. [investor-stated · strong] Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s” Observed results combine company development and market timing. [researcher inference · strong] Exit-market model
6. Global model failures. [researcher inference · strong] Japan, INF, ChinaVest, and Hong Kong weaken the idea that a local network plus Boston knowledge was reliably portable. Failures
7. Contradictory rhetoric and behavior. [researcher inference · moderate] Humble enabler versus active intervener; diversification versus follow-on faith; local autonomy versus central committee; stakeholder mission versus uncertain job/return evidence. Evolution
8. Memoir and memory risk. [researcher inference · strong] The 2009 book was collaboratively developed, partially accessible, and challenged by Cooksey; formation and succession dates conflict across later accounts. Winthrop project · CHM follow-up, printed pp. 59–62

The favorable thesis wins if ordinary, unseen decision files repeatedly connect Brooke-specific local information, vehicle design, candid-manager judgment, and governance actions to better outcomes after controlling for access, partners, stage, vintage, exit markets, and the top winner. The skeptical thesis wins if the missing files show generic selection, partner-led execution, weak ex-outlier economics, nonbinding governance, or returns driven primarily by market reopening and successor teams. [researcher inference · strong] Profile falsifiers · retrieval ledger

Analyst study guide, glossary, and go deeper

Recommended learning path

1. Read the final profile for chronology, people, vehicles, and the falsifiable edge.
2. Read the written corpus for direct arguments, access boundaries, and intellectual evolution.
3. Read the spoken corpus for Brooke's selection chain, admissions, and participant countervoices.
4. Work the nine no-hindsight decision labs before reading each actual outcome. Include the failure sequence Japan → INF → Hong Kong.
5. Use the mental-model field guide to turn the historical record into IC procedures, exceptions, and current transfer tests.
6. Finish with the evidence boundary; do not fill archival gaps with intuition.

Investor-specific glossary

Reusable checklists

Sourcing

• [researcher inference · moderate] Name the comparative advantage and the dated information the channel uniquely supplied. Model 3
• [researcher inference · strong] Record introducer, source owner, relationship duration, conflicts, funnel counts, and outcome by source. Model 12
• [researcher inference · moderate] Require a local sponsor with written authority, economics, references, and post-investment responsibility. Model 5

Diligence and IC

• [researcher inference · moderate] Test market growth, defensibility, commercializing management, candor, current technical validation, and explicit dissent. CHM, printed pp. 22–24 and 36–40
• [researcher inference · strong] Match instrument to cash stage; model company failure separately from security recovery. Model 2
• [researcher inference · moderate] Write thesis-kill criteria, evidence threshold, next-check milestone, and follow-on owner before approval. Model 13
• [researcher inference · moderate] Give the dissenting/local view an unedited section and record the committee response. Model 6

Portfolio and vehicle

• [researcher inference · strong] Keep legal entity, mandate, stage, geography, vintage, reserves, fees/carry, and cash flows separate. Model 12
• [researcher inference · strong] Show count and capital weights, losses, passes, results excluding the top winner, and exit-market assumptions. Model 9
• [researcher inference · moderate] Test partner workload, local/central authority, key-person triggers, succession, and committed support. Models 5, 10, and 11

Governance and value creation

• [researcher inference · strong] Pre-register owner, right, baseline, cost, milestone, time, downside, and counterfactual for every claimed intervention. Models 7 and 8
• [researcher inference · strong] Distinguish advice, board vote, management execution, and outcome; do not convert “we” into personal credit. Model 12
• [researcher inference · moderate] Surface bad news early, cap rescue capital, and document the stop or restructure decision. Models 6 and 13
• [researcher inference · moderate] Report financial returns and mission/stakeholder outcomes on separate scorecards. Candidate-model boundary

Ranked sources

1. CHM/NVCA Brooke oral history, printed pp. 1–63 — fullest direct account of Brooke's decision chain, attribution admissions, failures, and contradictions. [researcher inference · strong]
2. Concord oral history, paragraphs beginning “The department that I started,” “It seemed to me,” “Going back a little bit,” and “It wasn’t until the late ‘70s” — earlier direct account of bank sourcing, comparative advantage, and the exit drought. [researcher inference · strong]
3. 1975 Commercial and Financial Chronicle, printed/PDF pp. 5–7 — rare contemporaneous capital-structure statement with an immediate peer counterargument. [researcher inference · strong]
4. A Vision for Venture Capital, visible printed pp. 9–27, 65–127, and 166–193 — broadest operating synthesis, but partial access, collaborative production, and witness disputes require caution. [researcher inference · moderate] CHM follow-up, printed pp. 59–62
5. Florida/Kenney, printed pp. 114–123 — interview-derived bank/global mechanism and Aspen case, with author superlatives and source-mix limits. [researcher inference · moderate]
6. WBUR 2007, Brooke 10:00–11:16 and 22:44–24:10 and WBUR 2009, Brooke 06:55–11:28 and 27:23–28:51 — exact-audio controls on operating value, leverage, stakeholders, tax, and regulation. [researcher inference · strong]
7. Anagnostopoulos, 01:28:00–01:38:00, Tang, printed pp. 69–72, French, printed pp. 59–61, and Morby, 15:15–17:28 and 32:59–35:18 — essential participant controls against lone-founder attribution. [researcher inference · strong]

The highest-value missing sources remain the bank policy/loan files, the 1971 and 1973 article bodies, both Done Deals chapters, continuous access to A Vision, and TA/Advent agreements, IC materials, LP reports, board records, passes, write-offs, and cash flows. [researcher inference · strong] Retrieval ledger

Pareto 80/20 — what drove the record

Researcher inference · strong. No defensible lifetime 80/20 percentage can be calculated from the public record. Brooke moved through a bank loan book, Bessemer family capital, Tucker Anthony private placements, several TA partnerships and SBICs, Advent management companies, corporate accounts, country affiliates, and institutional funds. The available figures mix loan balances, company outcomes, a single-source attributed deal-multiple estimate, manager-reported whole-fund returns, and firm-wide assets. They cannot be added into a personal track record. Brooke oral history, printed pp. 16–47 · vehicle controls

Economic contribution — ranked within evidence class, not across unlike measures

Researcher inference · strong. No named Brooke-associated deal can be ranked by economic contribution from the public record. Biogen appears only as an unranked sensitivity sentinel; GPE II and ESSF remain separately ranked whole-fund controls. A remembered gross company estimate is not comparable with a fund's net IRR or multiple. Boston Globe obituary · Advent presentation, methodology and PDF p. 17 · reconciliation counts

Dominant-outlier tests. Researcher inference · strong. Removing the Biogen sentinel leaves zero named Brooke-associated deals with any public return-multiple claim. Removing either GPE II or ESSF from the separate whole-fund controls leaves only the other reported institutional observation; the canonical figures and their common manager-source boundary appear in the fund control. That sensitivity cannot establish repeatable Brooke selection or a literal “20% of deals produced 80% of returns” result.

Historical importance — ranked independently of return

Decision-learning value — ranked for replayability

Evidence boundary and searched denominator

Scope and inclusion rules

• Vehicles stay separate. Researcher inference · strong. A bank loan, Bessemer family position, Tucker placement, numbered TA partnership, SBIC, Advent corporate account, country affiliate, institutional fund and Brooke-family successor investment are different capital accounts unless a source proves continuity. CHM, printed pp. 16–47 · 1975 notice, printed p. 42805 · vehicle ledger
• A company enters the discovery ledger only through a target-specific source. Researcher inference · strong. Inclusion requires Brooke direct testimony, named participant testimony, a Brooke/TA/Advent institutional record, or a filing/press record tied to an already authenticated claim. Current selected portfolio pages do not backfill a Brooke portfolio. CHM, printed pp. 27–47 · Morby exact MP3, 15:15–29:22 · source-map coverage assessment
• Company outcomes are not investor returns. Researcher inference · strong. Bankruptcy, IPO, spin-off or acquisition proves a company event; without ownership, follow-ons, distributions, fees and dates it does not establish proceeds or a multiple for Brooke or a vehicle. Unitrode 1995 proxy, ownership table · Advent presentation, PDF p. 17 · deal ledger
• Roles do not collapse. Researcher inference · strong. Sourcing, sponsorship, diligence, approval, negotiation, board work, recruitment, follow-on decisions and exit work are credited separately. Founder status is not blanket deal attribution. Lea, printed p. 77; Tape 4, 00:02:36–00:04:06 · Aspen filing, “Management” · people map
• Outcome labels remain literal. Researcher inference · strong. “Reported” accompanies every unaudited multiple or IRR. Paper value, gross deal multiple, net fund multiple, net IRR, DPI and firm AUM are never summed. Boston Globe obituary · Advent presentation, methodology and PDF p. 17

Reconciliation counts

Researcher inference · strong. Coverage of Brooke's true investment population is not computable. Seventeen is a fully searched public name set, not the number of companies financed across the bank book, TA partnerships, Advent accounts or affiliates. Six additional program/platform/failure controls—the Bank loan book, Japan, INF, ChinaVest, the Hong Kong concept, and two unnamed failed software startups treated as one anonymous family—are analyzed separately and are not silently converted into company rows. The absence of a name from this ledger is not evidence that it was not funded. CHM, printed pp. 16–47 · Morby exact MP3, 17:28–29:22 · Tang, printed pp. 69–72 · program controls

Reproducible search boundary

Researcher inference · strong. The bounded search covered activity described from Brooke's 1956 bank start through the January 1996 contemporaneous CEO-succession record, with later outcome and retrieval sources checked through 18 August 2026. Name variants were paired with Bank of Boston, Bessemer, Tucker Anthony, TA Associates, Advent, each candidate company, fund aliases, investment, director, proxy, prospectus, merger, bankruptcy, failure, write-off, litigation, sanction, and return. Repositories included CHM/NVCA, Concord, HBS Baker Library, SEC EDGAR, GovInfo/Federal Register, Companies House, IFC/World Bank, public-LP materials, firm sites and contemporary press. Source-map attempted-route ledger

Historical market, regulatory, and base-rate context

• 1950s credit versus equity. Contemporaneous record · strong. The 1958 Small Business Investment Act created an SBIC framework; it did not itself create Brooke's bank mandate or approve any named loan. Brooke's reported use of assigned federal contracts as collateral was a credit response to young companies with brains and contracts but few conventional assets. Investor-stated · moderate. 1958 statute · Concord transcript, paragraphs beginning “The department that I started” and “The policy paper that I put”
• 1970s realization drought. Investor-stated · strong. Brooke says TA worked roughly ten to eleven years before substantial gains because public markets were largely closed; he dates meaningful liquidity to the 1979–1981 reopening. Researcher inference · strong: operating progress and exit-market timing are separate causal inputs. Concord transcript
• Security design under illiquidity. Investor-stated · strong. In 1975 Brooke described combining notes with common equity when new-issue markets and institutional supply were weak, creating a contractual repayment route instead of relying only on an IPO. Researcher inference · moderate: debt can reduce exit dependence but can also increase company fragility. 1975 trade-press passage, printed/PDF p. 6
• Pension capital. Contemporaneous record · strong. The 1979 ERISA final rule evaluated prudence in the context of the portfolio and diversification; reducing it to “venture capital became legal” would overstate the rule. 44 Federal Register 37221
• Emerging-market base rate. Contemporaneous record · strong. The IFC document records reported: −5.0% aggregate IRR for 21 early funds (IFC Note 90, printed p. 2, 1996; observed June 1983) and later identifies management quality, fund size, legal structure, exit markets, and local acceptance as recurring constraints. The same report lists SEFINNOVA, SEAVI, and an Advent private-equity fund among IFC-backed market-entry vehicles, but it does not isolate Brooke-era Advent returns. IFC Note 90, printed pp. 1–4

Vehicles, economics, and incentive boundaries

Researcher inference · strong. The source corpus exposes 25 distinct named vehicle/mandate rows or families, all searched: Bank technology lending; Bessemer Securities; Tucker Anthony; Advent I; Advent II; Advent (Bermuda) II; Devonshire Capital; Advent III; Advent IV partnership; Advent IV Capital; Advent V; Sofinnova; Advent Ltd./UK; the first Japan fund/affiliate; SEAVIC/SEAVI; Advent International Corporation; the Nabisco account; Microsoft/other corporate accounts; International Network Fund; ESSF; GPE II; first LAPEF; ChinaVest; Brooke Family LP/BPEA; and Vectis I/II. Four have hard public capitalization or narrow fund-economics evidence sufficient for the claim made—Devonshire, Advent IV Capital, ESSF and GPE II; 21/25 remain retrieval-only for economics, and 0/25 supports a Brooke-personal return. The table groups legally related rows for readability, not economic aggregation. 1975 notice, printed p. 42805 · 1981 notice, printed p. 52267 · Advent presentation, PDF p. 17 · source-map vehicle boundary

Researcher inference · strong. The nine reported corporate accounts and 23-affiliate network contain unnamed entities, so the actual vehicle total is not computable. Onshore/offshore partnerships, SBICs, management companies, client accounts, country funds and affiliates are never added as one pool. Anagnostopoulos, 01:28–01:38 · Florida/Kenney pp. 120–121

Incentive lesson. Researcher inference · moderate. A reported $150,000 fee on a $6 million early pool could sustain a very small team but created pressure to raise more capital; later Brooke described retaining some economics inside Advent to finance the network and criticized compensation detached from LP outcomes. The agreement-level carry, vesting, fee offsets, clawbacks and partner allocations remain unavailable, so this is a candidate mechanism rather than a proven causal account. A Vision, pp. 24, 79, 166 and 187–188

Dimension-matched contemporary LP-fund benchmark — as of 2026-08-18

Researcher inference · strong. This is a contract-dimension comparison, not a claim that Brooke-era vehicles had modern terms or that all current venture funds follow one template. The contemporary column uses ILPA's current published best-practice principles and definitions; those are recommendations, not law or a measured market average. ILPA Principles 3.0, printed pp. 10–22 and glossary p. 41

People, capital, and influence map

Comprehensive discovered-deal ledger

Researcher inference · strong. The ledger contains the 17 named company/opportunity candidates plus the Bank program as a separately counted context row. “Win,” “pass,” or “failure” records only the witness's classification; it does not establish a Brooke decision or return. CHM, printed pp. 16–47 · Morby exact MP3, 17:28–29:22 · Tang, printed pp. 69–72 · reconciliation counts

Program, platform, and anonymous-failure controls outside the 17-company denominator

Decision labs

Decision lab 1: Bank technology lending — underwriting a contract-backed credit program

• Decision-date freeze — researcher reconstruction; investor-stated premises · moderate. Young New England technology companies had federal R&D contracts but little conventional collateral. Assume contract payments can be assigned, but the policy paper, approval thresholds, borrower ledger, recovery priority, and loss history are unavailable. Concord transcript, paragraphs beginning “The department that I started” and “The policy paper that I put”
• Reader exercise — researcher-created, no hindsight. Would you approve the program then? Decide using federal-contract receivables and young-company credit quality. What would kill it? Unassignable contracts, weak recovery priority, borrower concentration, or a loss rate above the bank's threshold. What ownership/terms fit the risk? Senior contract-backed credit with covenants; equity only through a separately approved vehicle. What evidence unlocks the next check? Repayment history, contract quality, milestone delivery, exposure-at-default and committee approval.
• Sourced reveal — actual decision and outcome. Investor-stated · provisional. Brooke says the bank adopted contract-assignment lending, offered more favorable credit, observed borrowers for one or two years, and later retained selected relationships. He describes mutually unreconciled $10m/$20m exposure variants; the later $15k-loss claim is unaudited, and no company-level denominator survives. CHM, printed pp. 16–21 · Florida/Kenney, printed p. 116
• Post-reveal replay rule. Researcher inference · moderate. Senior debt and contract collateral can buy learning time only when recoveries, transferred exposures, later equity, and committee votes remain visible; otherwise favorable terms can magnify adverse selection.

Decision lab 2: Wang — underwriting a growing computer borrower and later governance

• Decision-date freeze — researcher reconstruction; investor-stated premises · moderate. A growing computer company has government-related receivables and seeks working capital. Treat the 1959/60 credit decision separately from any later equity or governance decision; equity, formal authority, board votes, succession alternatives, and platform-transition plans are unavailable. CHM, printed pp. 27–30
• Reader exercise — researcher-created, no hindsight. Would you extend credit or later invest then? What would kill it? Customer concentration, unfinanceable working capital, founder dependence, or no credible platform succession. What ownership/terms fit the risk? Loan covenants first; any equity requires explicit governance and succession rights. What evidence unlocks the next check? Aged receivables, backlog, cash conversion, product roadmap and an agreed succession milestone.
• Sourced reveal — actual decision and outcome. Investor-stated · moderate; contemporaneous record · strong for bankruptcy. Brooke reports offering more credit, a lower rate, and removal of Dr. Wang's personal endorsement; the bank financed Wang, and Brooke later joined the board and pressed succession. Wang filed Chapter 11 in 1992. No Brooke-attributable return is known. CHM, printed pp. 27–32 · New York Times, 19 August 1992
• Post-reveal falsification test. Researcher inference · strong. Board minutes and strategy papers should show a timely, technically credible transition plan sponsored by Brooke and rejected for identifiable reasons. Without them, “succession failure caused collapse” remains a participant interpretation; a long relationship and active director do not establish control.

Decision lab 3: Unitrode — underwriting a semiconductor position and management needs

• Decision-date freeze — researcher reconstruction; investor-stated premises · moderate. Assume a semiconductor company has a defensible power-integrated-circuit position and a material management gap. The original terms, co-director identity, board vote, candidate slate, compensation, customer validation, and operating plan are unavailable. CHM, printed pp. 32–33
• Reader exercise — researcher-created, no hindsight. Would you invest then? What would kill it? No independent validation of product advantage, an unfixable management gap, or board inability to recruit. What ownership/terms fit the risk? Enough equity and board rights to fund and govern a leadership transition without assuming control. What evidence unlocks the next check? Customer design wins, unit economics, candidate references and post-hire operating milestones.
• Sourced reveal — actual decision and outcome. Investor-stated · moderate; contemporaneous record · strong for the observed holding and merger process. Brooke claims a 1961 investment and says he and another director recruited new management. The 1995 proxy records his 36,000-share beneficial position, less than 1%, including 20,000 exercisable options; a separate 1999 board process produced TI's stock-for-stock acquisition. Holding continuity and return remain unknown. CHM, printed pp. 32–33 · Unitrode 1995 proxy, “Election of Directors” and “Security Ownership of Certain Beneficial Owners and Management” tables · TI–Unitrode joint proxy/prospectus, “The Merger — Background” and “Reasons for the Merger,” printed pp. 25–34
• Post-reveal replay rule. Researcher inference · moderate. Write the management gap, decision owner, candidate evidence and post-hire milestones before acting; record every director's contribution and do not let a favorable exit retroactively prove the diagnosis.

Decision lab 4: Federal Express — underwriting a distressed follow-on

• Decision-date freeze — researcher reconstruction; investor-stated premises · moderate. A capital-intensive air-freight network is near failure, senior financing is uncertain, and a down round is proposed. The sponsor, dissent, updated evidence, runway, ownership gained, portfolio concentration, and stop condition are not supplied in the surviving TA record. CHM, printed pp. 38–40
• Reader exercise — researcher-created, no hindsight. Would you fund or follow on then? Decide from route economics, aircraft financing and runway. What would kill it? Unit economics that worsen with scale, unavailable aircraft debt, missed network milestones, or a financing gap beyond reserves. What ownership/terms fit the risk? Staged preferred equity/subordinated debt with milestone releases, information rights and concentration limits. What evidence unlocks the next check? Load factors, contribution margin by route, cash burn, debt availability and signed customer demand.
• Sourced reveal — actual decision and outcome. Retrospective witness · strong for Lea's role; contemporaneous record · strong for court-described financings. Lea says he was principal lead and spent five years on Federal Express; the court record describes the 1973–74 financings; Brooke credits Lea and says an unidentified TA colleague increased the final-round exposure. TA participated, but no source shows Brooke led or supplies TA cash flows. Lea oral history, printed p. 77; Tape 4, 00:02:36–00:04:06 · court chronology · CHM, printed pp. 38–40
• Post-reveal replay rule. Researcher inference · strong. Sponsor persistence plus reserves can preserve nonlinear upside when evidence improves, but it can magnify loss when faith substitutes for milestones; every follow-on memo should name sponsor, decision owner, dissent, new evidence, concentration, and stop condition.

Decision lab 5: Biogen — biotechnology decision under an absent underwriting record

• Decision-date freeze — researcher reconstruction · insufficient public record. Treat this as an unidentified early biotechnology opportunity. No public source in the bounded search establishes the science thesis, source, TA partnership, round, check, ownership, board role, follow-ons, regulatory plan, exit dates, or gross/net definition; do not use any later multiple in the decision.
• Reader exercise — researcher-created, no hindsight. Would you invest then? Assess the science, intellectual property, team and capital plan. What would kill it? Irreproducible biology, weak patent position, no development path, or financing needs beyond the vehicle. What ownership/terms fit the risk? Staged equity tied to scientific and clinical milestones with syndicate reserves. What evidence unlocks the next check? Replication data, patent review, development milestones, regulatory plan and a costed financing schedule.
• Sourced reveal — actual decision and outcome. Researcher inference · insufficient public record; investor-stated · provisional for the estimate. The actual TA approval and Brooke's role remain unestablished. A 1980 shareholder-letter reference identifies TA participation, and a filed biography identifies Landry's board role. The only return figure is reported: ~50x (Boston Globe obituary, 2020), single-source, unaudited and unranked. BHR study, printed p. 203 n. 71 / PDF p. 18 · Landry SEC biography, “Kevin G. Landry” paragraph
• Post-reveal lesson. Researcher inference · strong. The case supports only the possibility of an important outlier, not a reusable biotechnology method. The next required evidence is the original TA memo and cash-flow schedule, not a richer retrospective story.

Decision lab 6: Aspen Technology — underwriting international expansion

• Decision-date freeze — researcher reconstruction; retrospective witness · moderate. A founder recalls Aspen as dangerously short of cash in 1986; Brooke later described a U.S.-only software company of roughly $4m “volume” and a thesis to expand internationally. The legal vehicle, stake, product retention, margins, acquisition budget, and causal operating plan are unavailable. Aspen founder presentation, PDF p. 17 · Florida/Kenney, printed pp. 122–123
• Reader exercise — researcher-created, no hindsight. Would you invest then? Underwrite the U.S. software base and international thesis. What would kill it? Weak retention, no repeatable sales motion, acquisition dependence without integration capacity, or negative unit economics. What ownership/terms fit the risk? Growth equity with board information rights and milestone-based acquisition capital. What evidence unlocks the next check? Cohort retention, margins, pipeline conversion, acquisition budget and named operating owners.
• Sourced reveal — actual decision and outcome. Retrospective witness · moderate for founder financing; investor-stated · provisional for Brooke's operating account; contemporaneous record · strong for Brown's role. The founder presentation reports $2.6m from Advent in 1986; Brooke later dates an investment around 1988 and attributes a British acquisition, Asian introductions and hiring, and a European acquisition/downsizing to Advent. The prospectus places Douglas Brown on the board from 1986. “Volume” and “value” remain undefined; costs, organic/acquired split, stake, proceeds, and a causal return bridge are absent. Aspen founder presentation, PDF p. 17 · Florida/Kenney, printed pp. 122–123 · Aspen 424B4, “Management”
• Post-reveal replay rule. Researcher inference · moderate. Pre-register baseline revenue, geography, acquisition price, integration cost, management owner, milestone, and counterfactual before calling internationalization investor value-add; management execution, market growth, and acquired revenue remain alternatives.

Decision lab 7: Japan — testing a cross-border affiliate model

• Decision-date freeze — researcher reconstruction; investor-stated premises · moderate. The proposed model pairs a credible local champion and local knowledge with shared venture standards and institutional capital. The public record does not supply the legal vehicle, portfolio, invested capital, committee contract, incentive schedule, exit map, or Japanese participant account. A Vision, printed pp. 65–71
• Reader exercise — researcher-created, no hindsight. Would you launch then? What would kill it? No aligned manager, ambiguous vetoes, inadequate fund scale, weak exits or entrepreneur resistance. What ownership/terms fit the risk? A staged local vehicle with explicit economics, information rights, central/local authority and a termination gate. What evidence unlocks the next check? Signed manager commitments, pipeline quality, LP close, decision matrix and test investments.
• Sourced reveal — actual decision and outcome. Investor-stated · moderate. Brooke says Advent launched the relationship/fund and later ended it, calling Japan a venture-capital failure; the memoir attributes the end to incompatible diligence and governance. Vehicle, capital, company outcomes, returns, committee record, and local countervoice remain absent. CHM, printed pp. 46–47 · A Vision, printed pp. 97–98
• Post-reveal countercase and replay rule. Researcher inference · strong. Culture is one interested explanation, not an adjudicated cause. Before launch, specify local sourcing authority, central veto, information rights, ownership, incentives, minimum viable scale, portfolio limits, and stop/restructure gates.

Decision lab 8: International Network Fund — allocating risk in a global pool

• Decision-date freeze — researcher reconstruction; investor-stated premise · moderate. Advent positioned the 1987 fund as a global institutional platform. Assume different country, manager, stage, and exit risks must coexist in one pool; the public record supplies no audited cash flows, sleeve limits, reserve policy, or company bridge. Advent history, “1987” milestone
• Reader exercise — researcher-created, no hindsight. Would you commit then? Test whether a single global pool can absorb different market and stage risks. What would kill it? Immature managers, weak exits, uncontrolled early-stage concentration or no comparable reporting. What ownership/terms fit the risk? Stage/geography sleeves, reserve limits, common reporting and committee escalation rights. What evidence unlocks the next check? Pipeline by stage, manager track record, exit map, loss budget and portfolio-construction model.
• Sourced reveal — actual decision and outcome. Investor-stated · moderate. Advent says the fund was raised in 1987; Brooke's memoir reports weaker results and excessive early-stage risk. No audited cash flows or company bridge surfaced. Advent history, “1987” milestone · A Vision, printed p. 114
• Post-reveal base-rate control and replay rule. Contemporaneous record · strong; researcher inference · moderate. IFC later identified small funds, inexperienced managers, weak exits, and one-tier governance as recurring emerging-market problems, but did not identify INF as part of its observed early-fund set. Allocate explicit risk budgets by stage, geography, and manager maturity; keep learning capital separate from core return capital. IFC Note 90, printed pp. 2–4

Decision lab 9: Hong Kong fund concept — underwriting sponsor commitments

• Decision-date freeze — researcher reconstruction; retrospective witness · moderate. Tang describes a proposed technology-transfer fund that depends on Advent/TA support and co-investment. Treat sponsor attention, staffing, technical help, and follow-on capacity as unproven until documented; no agreement, authority matrix, budget, pipeline, or deployment record survives. Tang transcript, printed pp. 69–72
• Reader exercise — researcher-created, no hindsight. Would you launch then? What would kill it? Unowned support obligations, missing local capital, no investable pipeline or irreconcilable sponsor priorities. What ownership/terms fit the risk? A separately governed vehicle with committed budgets, co-investment rights and termination remedies. What evidence unlocks the next check? Signed support schedule, capital commitments, pipeline, decision rights and escalation path.
• Sourced reveal — actual decision and outcome. Retrospective witness · moderate. Tang says the expected backing did not materialize and he left; Advent/TA records or responses were not accessible. No public evidence shows deployed capital, so this is an organizational commitment failure or abandoned platform rather than a realized investment loss. Tang transcript, printed pp. 69–72
• Post-reveal replay rule. Researcher inference · strong. Put sponsor attention, co-investment capacity, staffing, technical help, named owners, budgets, response times, escalation, and termination rights into the underwriting record.

Reported fund-level controls

Documented behavior · strong that Advent presented these figures in 2022; investor-stated · provisional for the unaudited manager-reported performance. Its methodology says net IRR uses LP contributions and distributions net of fees, expenses and carry, and that current values follow Advent's valuation policy; it also says the information was not audited or independently verified. Researcher inference · strong: the table is useful institutional evidence but cannot validate Brooke's personal economics, an early TA record, or a named deal. Presentation front methodology and p. 17

Portfolio-level forensics

• Selection versus exposure. Researcher inference · strong. The public record contains memorable survivors and one unranked reported multiple estimate but no complete portfolio or loss schedule. Hit frequency, loss ratio, sector attribution and vintage contribution cannot be calculated. Denominator and search boundary
• Liquidity dependence. Investor-stated · strong. A decade-long realization drought followed by a 1979–1981 window means observed proceeds depended on capital markets as well as company building. Concord transcript
• Stage drift. Investor-stated · moderate. Brooke's later postmortem says excessive early-stage risk weakened INF; his broader Asian account favors expansion capital outside the few ecosystems where technology venture worked. A Vision pp. 76 and 114–117
• Geographic heterogeneity. Researcher inference · strong. Sofinnova, SEAVI, Japan, corporate accounts, ESSF and GPE II had different owners, stages and governance. A “global network” aggregate would erase the mechanisms the evidence is trying to test. Advent history · IFC Note 90
• Operating value versus leverage. Investor-stated · moderate. Brooke's writings prefer acquisitions, market development, management work and conservative free-cash-flow underwriting over leverage alone. Researcher inference · insufficient public record: no portfolio bridge quantifies which mechanism produced the reported fund outcomes. A Vision pp. 157–167
• Attribution concentration. Researcher inference · strong. Brooke designed institutions, but Federal Express belongs first to Lea and the TA follow-on decision maker; later TA sourcing to Landry/Morby/team; Aspen operating credit to Brown, management and others; global cases to local managers and committees. A founder-only record contradicts the institution-building thesis itself. Lea oral history, printed p. 77; Tape 4, 00:02:36–00:04:06 · Morby exact MP3, 15:15–17:28 · Aspen 424B4, “Management”

Failures, misses, and anti-portfolio

Researcher inference · strong. The negative record is materially less complete than the success record. Morby supplies a partial named pass/failure set, but no audited write-off schedule, complete pass log, reserve history or litigation-by-investment ledger surfaced. “No substantiated personal venture-related sanction found” is a bounded search result, not proof of absence. Source-map critical coverage

Evidence-backed decision rules

1. Use structure to buy learning time. — Researcher inference · moderate. Contract-backed credit can create an information option before equity risk, but only if defaults, collateral and later exposure remain visible. Bank lab
2. Map comparative advantage before financing. — Investor-stated · strong. Brooke repeatedly favors sectors and locations with real knowledge or capability advantages. Concord transcript
3. Separate sponsor conviction from institutional decision. — Researcher inference · strong. Federal Express shows why the person who sources, holds the syndicate, approves and follows on must be named separately. Lea oral history, printed p. 77; Tape 4, 00:02:36–00:04:06 · CHM pp. 38–40
4. Board activity is bounded help. — Researcher inference · strong. Unitrode supports management intervention; Wang shows that an active director may still fail to change succession or platform trajectory. Unitrode 1995 proxy, “Election of Directors” and “Security Ownership of Certain Beneficial Owners and Management” tables · Wang bankruptcy
5. Globalization is a governance design, not a geography count. — Researcher inference · strong. Local information, ownership, incentives, central veto and support commitments must fit together; Japan and Hong Kong are the controls. A Vision pp. 97–127 · Tang pp. 69–72
6. Risk-budget learning capital. — Researcher inference · moderate. INF's reported stage drift and IFC's poor early base rate argue for explicit loss limits by stage, manager maturity and exit-market depth. A Vision p. 114 · IFC Note 90, pp. 2–4
7. Pre-register operating value. — Researcher inference · moderate. Aspen's mechanisms are plausible only when baseline, owner, cost, milestone, timing and counterfactual are written before results. Florida/Kenney pp. 122–123 · Aspen 424B4, “Management”
8. Never infer personal returns from company fame or fund AUM. — Researcher inference · strong. Biogen, FedEx, Unitrode and later Advent scale each demonstrate a different denominator trap. Biogen control · Unitrode 1995 proxy, “Election of Directors” and “Security Ownership of Certain Beneficial Owners and Management” tables · RI presentation p. 17

Uncertainty and retrieval ledger

Researcher inference · strong. Most remaining gaps are archival or owner-controlled after the logged open-web, filing, fund, company and direct-voice routes were attempted. The public record supports a bounded dossier but not a computed personal track record. CHM oral history, printed pp. 16–47 · Morby exact MP3, 17:28–29:22 · Tang transcript, printed pp. 69–72 · Source-map attempted-route ledger

Bottom line

Researcher inference · strong. Brooke's most defensible investment achievement is institutional: he repeatedly changed the vehicle—credit program, family-office investing, pooled TA partnerships, and a cross-border Advent platform—to match a changing opportunity set. The surviving economics are directionally favorable but thin: one unranked, single-source Biogen multiple estimate and two manager-reported institutional funds. The strongest learning comes from the frictions: Wang limits board power; Federal Express limits founder credit; Japan and Hong Kong limit easy replication; INF limits unconstrained learning capital; and Aspen limits retrospective value-creation claims. A complete personal return, hit rate, or 80/20 distribution is not supported by the public record. CHM oral history, printed pp. 16–47 · Lea oral history, printed p. 77 and Tape 4, 00:02:36–00:04:06 · Morby exact MP3, 15:15–17:28 · Tang transcript, printed pp. 69–72 · Advent presentation, PDF p. 17 · Reconciliation counts · failures, misses, and anti-portfolio

Coverage and evidence base

Literal priority-source reconciliation

At the WRITINGS cutoff, the source map contained 68 priority rows: 32 ★★★ and 36 ★★. T005.4 added 16 DEALS-routed priority rows; the live map now has 84 priority rows: 41 ★★★ and 43 ★★, reconciling exactly as 8 WRITINGS + 11 TALKS + 39 DEALS + 26 PROFILE/SYNTH = 84. Routing identifies primary phase ownership; it does not claim that a gated body was read. A source can still constrain another phase—for example, the Brooke oral histories are TALKS records but also prove that two claimed papers remain missing, while the Winthrop and Devaux records constrain authorship without becoming Brooke writings.

The four ★ rows are outside the literal priority denominator. They remain mapped for later use: the Endeavor and McGuire-event recaps are contextual, the Advent ESG publication is later institutional memory, and the Institutional Investor pioneer profile is a chronology-error lead. None is needed to complete the written-record accounting.

The 8 WRITINGS-routed priority rows reconcile as 2 full + 1 partial + 5 inaccessible/metadata-only: the full targets are the 1975 article and bounded Financiers section; A Vision is partial; the 1971, 1973, 1981, and two Done Deals bodies are blocked. Two further self-attested artifact leads—the Bank paper and claimed French subsection—were discovered inside TALK-routed oral histories. They are included in the note-backed artifact denominator below, but do not inflate the written priority-row count.

Written-record denominator

This phase produced 10 source notes for 10 distinct artifact or control families. Six are authenticated or claimed Brooke-authored works; four are edited-interview, reported-quotation, or scholarly controls needed to avoid mistaking another person's framing for Brooke's prose. The access accounting is complete:

The Brooke-authored or Brooke-claimed-artifact candidate sub-denominator is 6 = 0 full + 1 partial + 5 blocked. The partial work is A Vision for Venture Capital, which credits Brooke with Daniel Penrice and discloses Winthrop research/development/writing assistance. The five blocked candidates are the Bank policy paper, 1971 article, 1973 article, claimed French study subsection, and 1981 CODIMEC contribution. The 1975 article contains a reported Brooke quotation rather than Brooke-authored prose; Done Deals is edited interview material; Florida and Kenney wrote the manuscript section.

Exact access boundaries for the blocked record

Researcher inference · moderate: the sweep found a thin but larger written footprint than the original source map suggested. Three newly authenticated Brooke-attributed publication records—1971, 1973, and 1981—remain body-blocked. No public LP-letter series, personal memo archive, investment-committee corpus, portfolio review series, or audited early-fund cash-flow record surfaced. The only remaining high-value open-body lead would require new access rather than more ordinary web searching. Source-map attempted-route and archive audit

What the surviving written record actually establishes

The strongest contemporaneous direct evidence is narrow. In June 1975, amid weak public exits and longer holding periods, Brooke told a reporter that TA was using notes plus common equity rather than straight equity, so operating cash might provide a repayment path. Investor-stated · strong for the attributed financing adaptation; researcher inference · moderate for its downside logic; researcher inference · insufficient public record for issuer terms, prevalence, defaults, or returns. Pete Bancroft's adjacent objection—that debt is ill-suited to cash-poor startups—makes company stage an explicit boundary. Commercial and Financial Chronicle, printed/PDF p. 6

The partial 2009 memoir supports a much broader, but retrospective and collaboratively produced, system: relationship-led sourcing; selection tilted toward revenue and survivability; accountable local champions; local diligence paired with cross-market comparison; a later shift from loosely coordinated affiliates toward stronger central investment-committee authority; active ownership and operating improvement; management and information quality; admission of failure; and concern that leverage or compensation can detach returns from value creation. Investor-stated · moderate for the visible claims; researcher inference · insufficient public record for consistent execution or performance. A Vision, printed pp. 25–27, 65, 102–116, 152–188 · authorship/access note

Florida and Kenney's accessible 1997 manuscript section, drawing heavily on 1987 and 1993 Brooke interviews, independently preserves the local-plus-global operating hypothesis and Brooke's own qualification that Advent still had to prove the network generated better analysis. It also exposes attribution problems: the authors supply the largest/first superlatives; Brooke supplies selected recollections; other figures come from Joseph Powell, George Kenney, and firm brochures. Retrospective witness · moderate for the Brooke-attributed mechanism; researcher inference · strong that this is not independent corroboration of claims repeated from Brooke interviews. Financiers of Innovation, printed/PDF pp. 114–123 and endnotes 154–168

Overall boundary — researcher inference · strong: the surviving written record supports a stage-sensitive financing instinct, local-information thesis, evolving governance model, operating-value preference, transparency norm, team-development logic, and a mission larger than financial return. It does not establish Brooke's full sourcing funnel, valuation method, ownership targets, position sizing, reserve policy, board cadence, exit rules, person-level deal attribution, or audited performance across Bank of Boston, Bessemer, TA, Advent, country funds, corporate accounts, SEAVI, or BPEA. 1975 article, printed/PDF p. 6 · A Vision note · Financiers note

Source-by-source analysis

John Thackray — “The venture capital industry's troubled psyche” (1975)

Context, authorship, and access. The complete John Thackray article is visible in a FRASER scan of the June 16, 1975 Commercial and Financial Chronicle, printed/PDF pp. 5–7. Brooke is a quoted TA participant, not the author. The article combines market reporting, a price table, and comments from several venture investors. Full article, printed/PDF pp. 5–7 · source note

Central argument and reasoning. Thackray describes a venture market in which the new-issues outlet had weakened, institutional supply had receded, holdings lasted longer, and portfolio companies consumed more follow-on capital. Brooke's reported response was to combine notes and common equity, allowing company operations rather than an IPO alone to return capital. Investor-stated · strong for Brooke's attributed choice; researcher inference · moderate for the reconstructed chain: poor exit liquidity increases duration, duration increases financing exposure, and a contractual claim may create an alternate repayment path. Brooke passage, printed/PDF p. 6

Evidence, assumptions, and operational rule. The article gives no named Brooke transaction, security terms, maturity, covenant, seniority, cash-flow forecast, or realized result. Bancroft's immediate objection supplies the key assumption: debt works only if the company can service it without starving growth. Researcher inference · moderate: the transferable historical rule is not “use debt in venture”; it is “match repayment claims to stage and conservative operating cash.” Contrasting Brooke/Bancroft passages, printed/PDF p. 6

Actual decision, biography, and contradiction. The article identifies Brooke with TA in June 1975 but does not identify which partnership supplied capital or who approved a policy. The memoir's later warning that leverage alone cannot reproduce returns may be evolution, a distinction between moderate structured growth capital and buyout leverage, or simply different market contexts. The public record cannot choose among those explanations. Researcher inference · moderate. 1975 article, printed/PDF p. 6 · A Vision, printed p. 167

Brooke/Penrice — A Vision for Venture Capital (2009)

Context, authorship, and access. The title page credits Peter A. Brooke with Daniel Penrice; Brooke holds copyright, and the acknowledgments plus Winthrop's project page disclose research, development, interviews, and writing assistance. This is the developed Brooke/Penrice account, not a raw transcript or a basis for attributing every sentence solely to Brooke. Access was partial: front matter, contents, and discontinuous indexed snippets, not continuous chapters. Contemporaneous record · strong for authorship metadata; researcher inference · strong for the access boundary. Title/copyright, printed pp. iii–iv · acknowledgments, printed pp. 223–224 · Winthrop project

Publication context. The book appeared in 2009 after a credit cycle in which a 2008 GAO review had already highlighted leverage, underwriting, transparency, and oversight risks in large buyouts. The visible book's leverage critique fits that setting, but proximity does not prove the crisis caused Brooke's view or that his own earlier vehicles had the same exposure. Researcher inference · moderate. GAO-08-885, highlights and report · A Vision, printed p. 167

Central argument and reasoning chain. The visible record argues that venture and private-equity capability can contribute to economic development when locally informed investors combine capital, ideas, management development, governance, and international comparison. The Marshall Plan is an analogy, not the same institutional mechanism, and the authors concede the globalization project was only partly realized. Investor-stated · moderate for the mission and self-qualification; researcher inference · insufficient public record for causal or distributional proof. Introduction, printed pp. xiv–xvi · self-assessment, printed p. 146

Sourcing, picking, and portfolio logic. Relationships built before a financing need and credible local champions are the strongest visible sourcing signals. At TA, the authors describe most candidates as revenue-generating and profitable or near profitability, combined with diversification and investment protection. That supports a survivability-oriented growth model, not an audited claim that it outperformed early-stage power-law portfolios. Investor-stated · moderate. Printed pp. 25–27 · printed p. 26

Governance and information. The international narrative favors local knowledge but records a shift from affiliates operating in their own ways to a more Boston-centered committee that combined home-office and local analysis. Later chapters emphasize active ownership, apprenticeship across diligence/portfolio/exits/fundraising, full investment-committee information, and admitting defeat. Investor-stated · moderate for the visible model; researcher inference · strong that “local” and “central” are complements only when decision rights and escalation rules are explicit. Printed pp. 102, 110, and 113 · printed pp. 176–182

Value creation, incentives, and evidence. The visible cases favor operational improvement, conservative free-cash-flow underwriting, close monitoring, and management/environment/governance diligence over returns attributed to leverage alone. The authors also criticize manager wealth that can detach from fund performance and describe Brooke sacrificing personal carry allocations to build Advent. These are selected and self-reported; no partnership agreements or net cash flows test the incentive claims. Investor-stated · moderate. Printed pp. 116 and 157–167 · printed pp. 166 and 187–188

Actual decisions, biography, and contradictions. The book identifies an early $6 million Advent-named pool, a shared European management-company ownership design, later centralized Advent economics, Japan and ChinaVest disappointments, and Brooke's failure to create teamwork at TA. But its dates conflict with other records, its family tree is lineage rather than common ownership, and David Cooksey later disputed the account of his departure. Investor-stated · provisional for vehicle figures and selected explanations; retrospective witness · strong that an unresolved participant dispute exists. Printed pp. 22, 54, 78–80, 97–127, and 193 · 2010 Brooke oral-history addendum, printed pp. 60–61

Florida/Kenney — Financiers of Innovation, Brooke section (1997 manuscript)

Context, authorship, and access. The host exposes a complete 372-page February 1997 manuscript with draft placeholders and a “forthcoming” Princeton notice; no final published edition was verified. This phase analyzed the complete Brooke target at printed/PDF pp. 114–123 and its endnotes 154–168 on pp. 354–355, stopping where the next section begins. Florida and Kenney authored the prose; quotation marks and endnotes distinguish Brooke's 1987/1993 interviews from Powell, George Kenney, brochures, and author synthesis. Manuscript and front matter · target and notes · source note

Central argument and reasoning. Florida and Kenney use Brooke's career to tell a Boston-to-global institutional story: regional technology lending leads to TA's professional process, then to an Advent network that combines local market knowledge with international technology and competitor analysis. They supply the “first,” “most important,” and “world's largest” formulations; Brooke's attributed 1993 testimony explains the mechanism but says the network still had a long way to go and had to prove it produced better analysis. His Aspen example then names acquisitions, introductions, and local hiring as the claimed operating channels. Retrospective witness · moderate for the mechanism; researcher inference · strong against adopting the superlatives. Printed/PDF pp. 115 and 119–123

Evidence, assumptions, and operational rule. The Bank narrative assumes federal contracts can act as credit evidence; the TA narrative describes written qualification memoranda and three-person teams; the Advent narrative describes local investment committees, mixed diligence teams, and selected operating interventions. The pages expose no sample of rejected deals, comparative decision accuracy, affiliate control contracts, or net performance. Researcher inference · moderate: the defensible rule is to combine local context with cross-market comparison and documented team review, not to assume a global network creates an advantage. Printed/PDF pp. 115–123

Actual decisions, biography, and attribution. The source distinguishes Bank loans, another buyer's Wang equity, TA partnerships, Advent affiliates, institutional pools, and corporate separate accounts, but its narrative sometimes recombines them rhetorically. It reports $10 million of bank lending and $15,000 of loss from George Kenney's book, not a bank ledger; it calls Federal Express a Brooke hit although Brooke later credited Charlie Lea's syndicate leadership and another TA colleague's final-round decision. In the Aspen example, “we” leaves the legal investor and decision makers unnamed, while “volume” is undefined and the scale/geographic figures are unaudited. Researcher inference · strong that person, instrument, vehicle, and operating contribution must remain separate. Printed/PDF pp. 116–123 and notes 158–168 · Brooke oral history, printed pp. 39–40

Contradictions and missing witnesses. The manuscript itself gives TA start dates of 1966 and 1968 and Sofinnova dates of 1971 and 1972. It lacks Landry, Cooksey, local affiliate managers, LPs, failed-company leaders, and the underlying Brooke interview recordings. Those conflicts are useful evidence about source limitations, not minor editorial noise to average away. Researcher inference · strong. Printed/PDF pp. 117–119 and endnotes

Cross-source synthesis

Capital structure is stage-dependent, not ideological

The remembered Bank paper treats contracted research revenue as potential loan collateral; the 1975 article records note-plus-common structures during weak exit markets; the memoir later warns that leverage alone cannot reproduce attractive returns and favors conservative free-cash-flow underwriting. Researcher inference · moderate: the stable principle may be instrument fit—use repayment claims only when cash generation and downside terms can support them—rather than a move from pro-debt to anti-debt. The missing bank memo, transaction terms, and outcome ledger prevent a stronger conclusion. Bank-paper note · 1975 article, printed/PDF p. 6 · A Vision, printed pp. 116 and 167

Local knowledge requires shared standards and explicit authority

Both retrospective sources make the local champion central: local teams understand people, culture, regulation, and opportunity; the center contributes comparative technology and market knowledge. The memoir also records weak coordination and a later centralized investment committee. Researcher inference · strong: decentralization without common evidence standards can fragment judgment, while centralization without local veto or context can erase the claimed advantage. Neither source exposes the contracts and committee records needed to test the balance. A Vision, printed pp. 65, 71, 102–113 · Financiers, printed/PDF pp. 119–123

Operating help is a mechanism claim, not causal proof

The memoir selects operating improvement over leverage alone; the Florida/Kenney section records Brooke crediting Advent with acquisitions, Asian introductions, local hiring, and restructuring in Aspen Technology's expansion. Retrospective witness · provisional for the Aspen recollection; researcher inference · moderate that these are testable mechanisms; researcher inference · insufficient public record that they caused the reported scale or geographic mix. “We” leaves the investing vehicle and decision makers unidentified, while the section supplies no cost, margin, cash-flow, ownership, exit, or counterfactual ledger. A Vision, printed pp. 157–167 · Financiers, printed/PDF pp. 122–123

Bad-news transmission is a core governance mechanism

The memoir condemns withholding information from the investment committee and urges admitting defeat; Florida/Kenney's 1993 passage records Brooke qualifying his own network claim before proof existed. Researcher inference · moderate: the coherent operating norm is to preserve adverse evidence from local team through committee, define escalation, and stop when the thesis fails. The record does not show whether Brooke or his organizations applied that norm consistently. A Vision, printed pp. 177 and 182 · Financiers, printed/PDF p. 122

Institution-building and individual credit pull in opposite directions

The writings celebrate an institution that apprentices investors, reserves ownership for recruits, retains economics for expansion, and uses teams. Yet retrospective narratives repeatedly center Brooke and compress the bank, Bessemer, TA, Advent, affiliates, and later firms; the Aspen example says “we” without identifying the vehicle or individual decisions. Researcher inference · strong: if the institution-building claim is true, founder-only deal attribution becomes less—not more—credible. Landry, Cooksey, local managers, investment committees, and legal vehicles must remain visible. A Vision, printed pp. 54, 79, 176–193 · Financiers, printed/PDF pp. 117–123

Mission is explicit; causal and return evidence is not

The memoir links venture capability to jobs, companies, and economic development; Florida/Kenney frame Advent as a global institutional innovation. The same corpus admits disappointing Japan and broader Asian outcomes, limited non-U.S. venture returns outside Israel, and only partial realization of the mission. Investor-stated · moderate for the stated purpose and admissions; researcher inference · insufficient public record for incremental jobs, local spillovers, distribution, or net risk-adjusted performance. A Vision, printed pp. 76, 97–127, 146, and 149 · Financiers, printed/PDF pp. 119–123

The public corpus describes judgment more richly than it proves action

Intellectual chronology, evolution, and contradictions

Five tensions should remain explicit:

1. Debt as protection versus debt as fragility. — Researcher inference · moderate. The 1975 structure seeks operating repayment; Bancroft says startups cannot repay; the memoir later warns about leverage. Stage and cash profile may reconcile the views, but no transaction series proves that interpretation. 1975 article, printed/PDF p. 6 · A Vision, printed p. 167
2. Local autonomy versus central authority. — Researcher inference · strong. The global thesis depends on local judgment; the later governance correction centralizes committee control. The correct balance is not recoverable without affiliate agreements and minutes. A Vision, printed pp. 65, 102–113
3. Mission versus fiduciary result. — Investor-stated · moderate; outcome evidence insufficient. Brooke maintains that international participation produces learning and development even when many investors may lose money. The writings provide no capital-at-risk budget or threshold at which learning ceases to justify loss. A Vision, printed pp. 76, 146, and 194
4. Team institution versus founder narrative. — Researcher inference · strong. Shared ownership, apprenticeship, committees, and the unidentified “we” in the Aspen example conflict with retrospective assignment of entire firms and portfolios to Brooke. This is an attribution problem, not a reason to erase his role. A Vision, printed pp. 54 and 176–193 · Financiers, printed/PDF pp. 117–123
5. Authored doctrine versus mediated memory. — Researcher inference · strong. The richest visible work is coauthored and professionally developed; the scholarly section relies on interviews; the direct 1971/1973/1981 bodies are blocked. Confidence must therefore attach claim by claim, not to “Brooke's philosophy” as a single primary document. A Vision authorship note · Financiers authorship note · blocked-record table

Decision playbook reconstructed from the written record

This is an analyst reconstruction, not a Brooke-authored checklist. Each rule preserves its evidence grade and its failure boundary.

1. Match instrument to cash-generation stage. — Researcher inference · moderate. Ask whether conservative operating cash can service principal, interest, covenants, and growth needs before adding a repayment claim. Failure mode: importing a note-plus-equity structure into a pre-revenue company because it sounds protective. 1975 article, printed/PDF p. 6

2. Underwrite company economics separately from security protection. — Researcher inference · moderate. A senior claim can improve recovery while weakening the underlying company; record both enterprise downside and instrument downside. Failure mode: mistaking contractual priority for low business risk. Bank-paper note · A Vision, printed p. 116

3. Enter a geography through an accountable local champion. — Investor-stated · moderate. Ask who has local reputation, sourcing reach, operating judgment, economics, and decision responsibility. Failure mode: using a nominal affiliate or expatriate team as a substitute for locally embedded capability. A Vision, printed pp. 65 and 71

4. Pair local context with cross-market comparison and written decision rights. — Researcher inference · moderate. Require local and central analyses, define veto/escalation authority, and record disagreements. Failure mode: either central override without context or decentralized inconsistency without shared standards. A Vision, printed pp. 102–113 · Financiers, printed/PDF pp. 121–123

5. Make negative information travel intact. — Investor-stated · moderate. Put disconfirming facts, missing data, and local objections into the same committee packet as the upside case. Failure mode: filtering evidence to what a sponsor thinks the committee needs. A Vision, printed pp. 177 and 182

6. Underwrite and train the people who will own the whole job. — Investor-stated · moderate. Evaluate management, then apprentice investors across diligence, portfolio work, exits, and fundraising rather than rewarding isolated origination. Failure mode: star culture that hides weak teamwork or creates key-person dependence. A Vision, printed pp. 152 and 176–193

7. Name the operating value-creation mechanism before relying on leverage. — Investor-stated · moderate. Specify product, pricing, route-to-market, management, governance, or cost changes and the owner responsible. Failure mode: accepting a selected acquisition-and-expansion story as causal proof without identifying vehicle, costs, management contribution, or counterfactual. A Vision, printed pp. 157–167 · Financiers, printed/PDF pp. 122–123

8. Separate mandates and risk budgets by stage. — Researcher inference · moderate. Keep early venture, growth, restructuring, corporate strategic accounts, and country funds distinct; measure concentration and reserves within each. Failure mode: aggregating unlike pools into one global success narrative. A Vision, printed pp. 26 and 114–117 · Financiers, printed/PDF pp. 120–121

9. Align personal economics with fund and institution outcomes. — Investor-stated · moderate. Trace management fees, carry, retained corporate economics, ownership, and forfeiture through the actual agreements. Failure mode: accepting a retrospective sacrifice or alignment story without allocation schedules and LP cash flows. A Vision, printed pp. 166 and 187–188

10. Predefine defeat and run a falsifiable postmortem. — Researcher inference · moderate. State what evidence kills the thesis, who can stop follow-ons, and how failures update geography, stage, team, and governance choices. Failure mode: converting every loss into an unfalsifiable “learning investment.” A Vision, printed pp. 97–127 and 177

11. Ledger person, vehicle, instrument, and decision separately. — Researcher inference · strong. For each check, record legal investor, partnership, approvers, syndicate lead, board representative, security, ownership, follow-ons, and exit. Failure mode: assigning Bank capital, a team decision, an affiliate deal, or an unidentified “we” result to Brooke personally. Financiers, printed/PDF pp. 116–123

12. Measure mission and financial performance on separate scorecards. — Researcher inference · strong. Track jobs, capability transfer, local ownership, company outcomes, gross/net cash flows, risk, and benchmark separately. Failure mode: using economic-development rhetoric to imply fund performance, or weak returns to erase genuine capability building. A Vision, printed pp. 76, 146, 149, and 194

Modern VC translation (as of 2026-08-02)

Contemporaneous record · moderate: the Q2 2026 PitchBook-NVCA Venture Monitor says first-half U.S. startup investment exceeded $400 billion, while AI, mega-rounds, established managers, and a relatively small set of exits dominated the recovery. The 2026 NVCA Yearbook reports that in 2025, 487 $100 million-plus rounds were 3.2% of deal count but 67% of deal value, and two-thirds of unicorn IPOs priced below their last private valuation. These association/data-provider observations establish current concentration and uneven liquidity; they do not validate a Brooke rule or predict any company. Q2 2026 Venture Monitor · 2026 NVCA Yearbook

The blocked 1973 pension article contributes no modern pension rule. Current U.S. ERISA investment duties require plan fiduciaries to evaluate portfolio role, risk/return relative to alternatives, diversification, liquidity, current return, projected return, and funding objectives; a title from 1973 cannot replace that current analysis. Contemporaneous record · strong. 29 CFR §2550.404a-1(b)

Each of the 12 reconstructed historical rules maps once below. Current examples that are not directly evidenced are explicitly labeled hypotheses.

Five writings to read or retrieve first

Ranking method — researcher inference · moderate: prioritize contemporaneity, Brooke attribution, decision usefulness, adversarial value, access completeness, and ability to change the synthesis. This is a research order, not a performance or literary ranking.

1. “The venture capital industry's troubled psyche” (1975) — available in full. It is the clearest contemporaneous Brooke decision signal and includes Bancroft's immediate stage-based counterargument. Full scan, printed/PDF pp. 5–7 · note
2. A Vision for Venture Capital (2009) — only partially available in this run. It is the broadest operating and intellectual synthesis, but must be read as Brooke/Penrice/Winthrop-developed retrospective work and checked against participant disputes and outcomes. Google Books record · note
3. Financiers of Innovation, Brooke section (1997 manuscript) — target available in full. It preserves the 1987/1993 interview-based local/global mechanism and Aspen operating example and, crucially, separates Brooke testimony from author superlatives and mixed-source figures. Target, printed/PDF pp. 114–123 · note
4. “How a Private Venture Capital Firm Operates” (1971) — retrieve; body inaccessible. This four-page sole-authored article is the highest-value missing direct work because it may reveal an early operating doctrine before later institutional hindsight. JSTOR record · route note
5. “Should Pension Funds Participate in Venture Capital?” (1973) — retrieve; body inaccessible. It may precede and clarify later pension-capital advocacy, but the title currently supports no answer or policy claim. Official bibliography, printed p. 75/PDF p. 85 · route note

The 1981 CODIMEC contribution ranks immediately after these five because it may reveal how Brooke presented TA to a non-U.S. audience before the TA–Advent separation. Its form and body are less securely established than the 1971 and 1973 records. CODIMEC route note

Unresolved written record and retrieval agenda

Planning label — researcher inference · moderate: priority reflects expected ability to change the synthesis, not certainty that the named repository holds the record.

1. Retrieve the 1971 article. Use JSTOR, HeinOnline, the ABA archive, interlibrary loan, or a physical Business Lawyer 26(3), pp. 735–738. Capture role line, examples, rules, and every page. Route note
2. Resolve and retrieve the 1973 article. Identify the exact Pensions serial, publisher, issue, and pages through union catalogs and pension-industry archives; do not normalize it to Pensions & Investments without evidence. Route note
3. Recover the Bank policy paper and decision trail. Begin with a targeted request to Bank of America Corporate Archives' restricted FleetBoston/acquired-institution holdings, then search Bank of Boston/First National Bank records, Bill Ray and Bill Brown papers, executive and credit-committee minutes, borrower files, and Brooke family materials. Reconcile the reported $10 million and $20 million portfolio figures with an actual ledger. Bank-paper note
4. Resolve the French artifact family. Begin with MIT's Arthur D. Little records MC-0579, then inspect SAEF 1A127/3, Archives nationales 19820043/1, the full Le Cahier de l'Innovation, and ministry contracts. Determine whether Brooke supplied a separate memo, interview, appendix, or translated section rather than assuming authorship of the white paper. French-study note
5. Locate the 1981 CODIMEC proceedings. Search CODIMEC/CODEMEC, ACRJ, FIESP, CVM, FGV, Biblioteca Nacional, and participant archives for the program and white paper; establish language, editor, translator, pages, and Brooke review. CODIMEC note
6. Borrow both Done Deals chapters lawfully. Process Brooke pp. 245–258 and Landry pp. 259–268 together, recording interview date, editor/voice boundaries, vehicle chronology, partner attribution, claims, and contradictions. Brooke route · Landry route
7. Obtain continuous access to A Vision. Read all 275 pages rather than extending discontinuous snippets; reconcile every case, date, failure, partner claim, ownership statement, and reported return with external records. Partial-access note
8. Recover the absent operating record. TA/Advent partnership agreements, investment memoranda, committee minutes, LP letters, board records, portfolio reviews, passes, write-offs, exits, and cash flows are required to test whether the written principles governed decisions and produced results. Source-map denominator audit

Researcher inference · strong: until those records are recovered, the sharpest defensible conclusion is bounded. Brooke's surviving written record describes how stage, instrument, local knowledge, comparative judgment, governance, people, incentives, failure, and economic mission might interact. It does not prove a complete personal investment system or its returns. 1975 trade-press scan, printed/PDF pp. 5–7 · Financiers of Innovation, printed pp. 114–123 · A Vision partial-access note · written-record denominator · retrieval agenda

Coverage and evidence base

The TALKS corpus contains 11 deduplicated priority families, all represented by one source note: 6 full oral-history transcripts, 3 full exact-file audio appearances, 1 full exact-file anniversary video with captions, and 1 forum family with partial contemporaneous reporting but no authenticated stage audio or transcript. Contemporaneous record · strong for the inventory and access states; researcher inference · strong for the family boundaries. Source map

The WBUR and Morby files are bound by canonical page, direct asset URL, retrieval date, bytes, runtime, and SHA-256. OpenAI API transcription was used only as a locator aid; material speaker identities were normalized from introductions, official guest lists, continuous turns, and context. No raw transcript, audio, caption, or API response is stored in the repository. The 2009 WBUR page displays 24:22 while its live Megaphone asset runs 46:41.946; every locator therefore binds to the exact fingerprinted MP3, not the page display. WBUR 2007 note · WBUR 2009 note · Morby note

Four neighboring source-map records are controls rather than appearances: the NVCA collection announcement, Wang finding aid, Winthrop memoir page, and HBS award register. The 1975 trade-press quotation, two Done Deals interview chapters, and interview-derived Financiers of Innovation material remain in the written corpus and are not double-counted. Researcher inference · strong. Source map

Source-by-source guide

Peter Brooke — Concord Oral History Program (1994)

Brooke links contract-backed technology lending, observation before equity investment, comparative advantage, the 1970s realization drought, and international affiliates. He says TA worked roughly ten or eleven years before substantial gains, separating company development from exit-market timing; he also uses both 1967 and 1971 around TA, preserving rather than resolving a formation-date conflict. Investor-stated · strong for his 1994 recollection; researcher inference · insufficient public record for legal dates, cash flows, and causal performance. Concord transcript, paragraphs beginning “The department that I started,” “Going back a little bit,” and “My firm TA Associates began” · note

His most decision-useful chain is stage-specific: lend against observable contracts, learn through the customer relationship, and finance where a region has genuine knowledge advantage. The transcript supplies no underwriting thresholds, rejected-company denominator, losses, or ownership schedule. Investor-stated · moderate. Concord transcript, paragraphs beginning “The policy paper that I put” and “It seemed to me”

Peter Brooke — CHM/NVCA oral history (2010)

The 69-page interview is the corpus anchor. Brooke distinguishes selecting risk from enjoying risk; calls entrepreneurs the protagonists and investors enablers; and describes a five-part pattern: growth sector, defensible technology, commercializing management, current technical expertise, and outside dissent. Investor-stated · strong for the articulated framework; researcher inference · provisional for consistent application or efficacy. CHM transcript, printed pp. 6, 11–12, 22–24, and 36–40 · note

The same source limits Brooke-wide credit and borderless-network mythology. He assigns Federal Express to Charlie Lea, distinguishes Bank lending from later board and TA roles, credits Landry in the TA–Advent split, says overseas early-stage technology worked only in Israel, and calls Japan a failure. The follow-up records David Cooksey's objection without an independent Cooksey account or recording. Investor-stated · strong for Brooke's admissions and attribution; researcher inference · insufficient public record for legal decision rights, full losses, and disputed history. CHM transcript, printed pp. 27–47 and 59–62

T. Bondurant French — CHM/NVCA oral history (2018)

French assigns the later systematic TA model to Kevin Landry and a partner team, discusses spinouts and the International Network Fund, and separates TA's domestic continuation from Brooke's Advent path. Retrospective witness · moderate: this constrains founder-wide credit but does not replace partnership, allocation, committee, or fund records. French transcript, printed pp. 59–61 · note

Constantine Anagnostopoulos — Science History Institute oral history (2012)

Anagnostopoulos says he helped build Advent UK with corporate and university-linked capital, recruited David Cooksey, proposed a corporate-account model, and held cofounder/director roles. He also places a Genzyme-related decision around 1982, before official Advent International chronology. Retrospective witness · moderate: the participant account makes a Brooke-only origin story untenable, while incorporation, subscription, board, and vehicle records remain necessary to allocate legal credit. Synchronized transcript/audio, 01:28:00–01:38:00 and 02:00:48–02:04:00 · note

Martin Tang — MIT oral history (2021–2022)

Tang recalls a Hong Kong technology-transfer vehicle that did not receive the Advent/TA support or co-investment he expected. Retrospective witness · moderate: this is material negative evidence against a frictionless local-plus-global network, but requested capital, entity boundary, committee reasoning, and financial outcome are unavailable. MIT transcript, printed pp. 69–72 · note

Paul Bancroft III — Bancroft Library oral history (2010)

Bancroft narrows Brooke's pre-Bancroft Bessemer role, keeps Charlie Lea visible, and places both inside a wider early venture network. Retrospective witness · moderate: this is useful credit-assignment evidence, not a Bessemer deal ledger or return history. Bancroft transcript, printed pp. 45 and 59 · note

WBUR — Big Deals in Private Equity (2007)

Brooke argues that legitimate private equity improves operations rather than merely adding leverage; warns that cheap debt, rate changes, and bank liquidity can create danger; condemns asset stripping; and proposes union dialogue during restructuring. He nevertheless defends a majority of managers without a cited denominator and asks for job/value study before carried-interest reform. Investor-stated · strong for his positions; researcher inference · insufficient public record for prevalence, causality, or policy effect. Exact 43:43.216 MP3, Brooke, 09:36–15:53, 18:56–24:10, and 33:19–38:32 · note

The countervoices are central evidence, not scenery. Stephen Lerner challenges opacity, debt-loaded nondistressed service buyouts, and worker distribution; Colin Blaydon separates purchase price, leverage, and cash-flow improvement; callers describe human costs; Roben Farzad frames cheap debt, layoffs, fees, and the industry's operating-change defense. Contemporaneous record · strong for their stated arguments; researcher inference · insufficient public record for unnamed deals and aggregate effects. Exact MP3, 01:26–08:28, 16:34–18:40, 20:34–22:04, 24:13–36:35

Brooke–Terry McGuire Xconomy forum family (2009)

A recovered September 24 Xconomy article says reporter Robert Buderi spoke with Brooke the previous day. Brooke's reporter-attributed positions favor constructive company-building, expect some regulation, warn against a response driven by bad actors, and distinguish private-equity efficiency from venture creation. This is a separate September 23 reporter contact, not October 21 stage testimony. Investor-stated · moderate. Recovered source coordinates and route ledger

The October 20 preview quotes McGuire, the October 29 gallery confirms the event and promised a later writeup/video, and Steven Dickman's attendee account reports LP-allocation pressure. The promised authenticated stage package did not surface in the recorded public routes. Contemporaneous record · strong for occurrence and publication plan; contemporaneous record · provisional for the attendee account; researcher inference · insufficient public record for Brooke's complete stage remarks. Route note

WBUR — Peter Brooke: Private Equity Now (2009)

Brooke separates equity-funded venture from leveraged buyouts, claims operating value and job creation, admits early layoffs and destructive ownership, condemns Simmons, and acknowledges a large debt overhang. His job-growth and majority-good-actor claims have no supplied dataset. Investor-stated · strong for the distinctions and admissions; researcher inference · insufficient public record for quantities and prevalence. Exact 46:41.946 MP3, Brooke, 02:29–11:28 · note

His regulatory stance is modular: accept SEC registration/reporting, resist regulators choosing investment risks, consider narrow activity boundaries when PE crosses into banks or hedge funds, concede Fleischer's tax-fairness point, and resist fixed leverage rules. Fleischer argues for systemic reporting and ordinary-income treatment of carry; Tan Keng Boon describes a regulated, equity-focused affiliated Asian model. Investor-stated · strong for Brooke's responses; contemporaneous record · strong for Fleischer's proposal; investor-stated · moderate for Tan's affiliated account; researcher inference · insufficient public record for efficacy. Exact MP3, 23:19–33:01 and 41:09–45:55

Advent International — 40th-anniversary witness video (2024)

Six people shown as Advent managing partners narrate the 2024 anniversary. James Brocklebank's continuous segment recalls Brooke describing Advent as a “noble experiment” built around locally rooted firms and economic development. This is Brocklebank's later recollection, not Brooke audio, formation proof, or evidence about every vehicle. Retrospective witness · moderate. LinkedIn video, Brocklebank, 00:28.640–00:48.890 · speaker/version note

Jacqui Morby — The Queen of Software (2025)

Morby remembers a roughly ten-person TA with Brooke running the firm and Kevin Landry heading investing. She says Brooke hired her and encouraged openness, while she devised and taught the outbound magazine/Dun & Bradstreet/cold-call workflow and brought leads to Landry; she explicitly says Brooke was not much involved in that system. Retrospective witness · strong for Morby's own process; retrospective witness · moderate for colleagues' roles. Exact 40:06.478 MP3, Morby, 08:46–10:00, 15:15–17:28, and 32:59–35:18 · note

Her software screen emphasized repeatable products, observable adoption, customers, and references; she also admits failed startup investments and named misses. These are Morby's later-TA rules, not Brooke or Advent rules, and the episode lacks the full opportunity/return denominator. Retrospective witness · strong for her stated process; researcher inference · insufficient public record for efficacy. Exact MP3, 23:53–29:22

Recurring frameworks and strongest counterarguments

Selection and staged information

Brooke consistently frames venture work as selection and staged exposure rather than appetite for danger: observe contract-backed borrowers, test management candor, add current outside expertise, and preserve dissent. Investor-stated · strong for the framework; researcher inference · provisional for actual use or results because declined-company, loss, and cash-flow denominators are missing. Concord transcript · CHM transcript, printed pp. 6 and 22–24

Operating value must be separated from leverage and distribution

The WBUR record provides a practical falsification test: separate entry price, debt, fee extraction, operating cash-flow improvement, product/R&D investment, employment, and exit-market beta. Brooke's operating-value defense survives only deal by deal; his own admissions about overleverage and asset stripping, Lerner's worker critique, Blaydon's return decomposition, and callers' accounts prevent a portfolio-wide inference. Researcher inference · strong. WBUR 2007 exact MP3, 09:36–15:53 and 16:34–36:35 · WBUR 2009 exact MP3, 03:58–11:28 and 20:22–38:06

Brooke's German union-dialogue example sits in tension with his later framing of pensions and employee benefits as a separate system question. The positions are not logically incompatible, but the category boundary weakens his broad stakeholder defense. Researcher inference · strong. WBUR 2007 exact MP3, 18:56–20:22 and 31:49–33:19

Institution-building is a team outcome

Brooke names Lea and Landry; French credits Landry and partners; Anagnostopoulos claims capital/model/board/recruiting work; Bancroft narrows the Bessemer role; Morby assigns herself the outbound system while distinguishing Brooke's culture from Landry's investing authority. Researcher inference · strong: the public testimony rejects a lone-genius account, though legal decision rights remain unknown. CHM transcript, printed pp. 38–43 · French transcript, printed pp. 59–61 · Anagnostopoulos, 01:28:00–01:38:00 · Morby exact MP3, 08:46–17:28

Local knowledge is not a substitute for explicit decision rights

The favorable theory combines local information with cross-market learning. Brooke's Japan admission, Tang's weak-support account, Anagnostopoulos's locally designed capital model, and Tan's affiliated equity-first description show that “global network” covered different strategies and governance arrangements. Researcher inference · strong for heterogeneity; researcher inference · insufficient public record for which design produced better net returns. CHM transcript, printed pp. 44–47 · Tang transcript, printed pp. 69–72 · WBUR 2009 exact MP3, 41:09–45:55

Exit markets are part of the return mechanism

Brooke's 1994 account separates company-building from realization: a long 1970s drought preceded stronger 1979–1981 public exits. Investor-stated · moderate for his sequence; researcher inference · insufficient public record for how much outcome came from selection, operations, leverage, price, or market reopening. Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s”

Intellectual evolution and unresolved contradictions

Researcher inference · strong: the regulation record is not a pro/anti binary. Reporting, systemic monitoring, regulator investment judgment, activity boundaries, leverage caps, carried-interest taxation, labor protections, and self-policing are distinct choices. Brooke changes position or emphasis across some dimensions while remaining consistently resistant to blunt, ex-ante rules. WBUR 2007 exact MP3, 22:44–24:10 and 36:35–38:32 · WBUR 2009 exact MP3, 23:19–38:06

The sharpest chronology conflicts remain unresolved: TA activity/formation appears as 1967, 1968, or 1971; Sofinnova as 1972 or 1973; Advent as 1984 or 1985; and the CEO handoff across 1995–1997. These require record-type reconciliation rather than majority vote. Researcher inference · strong. Concord transcript · CHM transcript, printed pp. 35–43 and 58 · Anagnostopoulos, 01:28:00–01:38:00

Biography, deals, vehicles, and attribution leads

Five sources to study first

1. Brooke CHM/NVCA oral history — richest direct record for process, deals, team culture, global limits, succession, and the Cooksey dispute.
2. Brooke Concord oral history — earlier memory checkpoint for contract-backed observation, comparative advantage, exit drought, and date tensions.
3. WBUR Big Deals in Private Equity — strongest contemporaneous adversarial test of operating value, leverage, labor, distribution, tax, and self-policing.
4. WBUR Peter Brooke: Private Equity Now — best evidence of Brooke's post-crisis leverage admissions and modular regulation/tax position.
5. Anagnostopoulos oral history — strongest capital/cofounder/country-fund counterweight to a Brooke-only Advent history.

Morby is the next attribution control because it cleanly separates Brooke's culture, Landry's investing authority, and Morby's operating system. French, Tang, and Bancroft provide compact team, failure, and apprenticeship controls.

Evidence still missing

• No authenticated audio, transcript, or promised full writeup for the October 21 Xconomy stage discussion surfaced; the separate reporter interview, preview, gallery, and attendee account cannot substitute for it. Exhausted route ledger
• Incorporation, board, partnership, subscription, asset-transfer, investment-committee, employment, and correspondence records are required to adjudicate Cooksey, Anagnostopoulos, Landry/team, Morby, and Tang accounts.
• No public corpus of Brooke speeches, LP communications, investment-committee recordings, annual-meeting remarks, or standalone testimony surfaced. The mapped 1971/1973/1981 writing-or-speech artifacts remain body-blocked in the written corpus.
• Vehicle cash flows, leverage, fees, reinvestment, employment, failed/unexited deals, and benchmark denominators are still required before treating Brooke's operating-value, job-creation, global-success, or majority-good-actor claims as measured results.

Researcher inference · strong: the spoken record supports a bounded portrait—selection, comparative advantage, local information, outside expertise, active but limited investor help, institution-building, and mission—alongside slow realizations, debt harms, Japan, overseas technology limits, labor/distribution tensions, and disputed founder credit. It does not prove a complete personal investment system, a uniform global model, founder-exclusive credit, or audited performance. CHM oral history, printed pp. 11, 22–24, and 44–47 · WBUR 2007 exact MP3, 09:36–15:53 and 31:49–38:32 · Morby exact MP3, 15:15–17:28 and 32:59–35:18 · coverage denominator · missing evidence

1. Use credit as an information option only when observation changes the next decision

• Rule — [researcher inference · moderate] A protected financing relationship can buy time and behavioral information before a higher-risk commitment, but only if the lender defines what it will learn, when it will reassess, and how it will cap credit exposure.
• Evidence status — [retrospective witness · moderate] Brooke twice described the bank program as attracting a broad technology-company set, observing borrowers, and retaining selected relationships; [researcher inference · insufficient public record] for its selection efficacy because the paper, borrower funnel, defaults, and later investment outcomes are missing. Concord transcript, paragraphs beginning “The department that I started” and “The policy paper that I put” · CHM transcript, printed pp. 16–21
• Provenance and origin — [investor-stated · moderate] The earliest supportable use is Brooke's remembered 1958-focused Bank of Boston program: lend against assigned federal contracts, observe companies for one or two years, and preserve relationships with the strongest. The internal paper is missing, so no wording is attributed to it. Concord transcript, paragraphs beginning “The department that I started” and “The policy paper that I put”
• Mechanism and reasoning — [researcher inference · moderate] A live credit relationship reveals repayment behavior, cash conversion, customer concentration, reporting quality, and management candor before equity selection; seniority limits loss distribution but does not remove company risk. Bank-paper note · Wang lab
• Historical and vehicle context — [investor-stated · moderate] Brooke recalled that young technical companies lacked conventional assets and that federal contracts could support assignment-based lending; the bank—not Brooke personally—owned the balance sheet and approval. [contemporaneous record · strong] The 1958 Small Business Investment Act created a separate SBIC framework and did not authorize the bank program. CHM transcript, printed pp. 16–21 · 1958 statute
• When to use — [researcher inference · moderate] Use when a company has verifiable contracted cash flow, the credit period will expose a decision-relevant behavior, and downside can be bounded. Do not use for pre-revenue science, structurally negative cash flow, or when lender covenants would suppress the very experiment being evaluated.
• How to apply it — [researcher inference · moderate] (1) define collateral and recovery; (2) list the three facts the relationship must reveal; (3) set reporting/covenants; (4) cap exposure and concentration; (5) pre-set the equity or renewal decision date; (6) require changed evidence rather than relationship familiarity. Bank decision lab
• Examples
  • Bank technology-loan book — [documented application, investor-stated · moderate]. Brooke described contract-backed credit as both financing and an observation funnel; the reported $10m/$20m/$15k figures remain unreconciled. CHM, printed pp. 16–21 · Florida/Kenney, printed pp. 115–116
  • Wang — [documented application, investor-stated · moderate]. The bank relationship generated access and later board involvement, but the subsequent succession/platform failure shows that accumulated information is not control. CHM, printed pp. 27–32 · NYT bankruptcy report
  • Biogen — researcher-applied boundary. [researcher inference · strong] A later single-source multiple estimate cannot substitute for an original learning path, scientific evidence, or staged decision record; no credit-first application is claimed. Biogen lab
• Evidence — [researcher inference · moderate] The mechanism has two separate Brooke accounts plus the Wang outcome control; it lacks contemporaneous bank files, a complete funnel, and an equity-conversion ledger. Concord transcript, paragraphs beginning “The department that I started” and “The policy paper that I put” · CHM, printed pp. 16–21 and 27–32 · HBS Wang finding aid, subseries IIB
• Limits, failure modes, and exception ledger — [researcher inference · strong] Relationship familiarity can masquerade as learning; collateral can create false confidence; a lender may learn but lack equity rights; and credit can damage a cash-poor company. Missing routes: Bank of America Corporate Archives/FleetBoston holdings, Bank executive and credit-committee minutes, borrower files, Bill Ray/Brown papers, and Wang investment records. Retrieval ledger
• Evolution and contradictions — [researcher inference · moderate] Brooke later moved from bank observation to private equity and pooled funds, but no evidence shows a formal credit-to-equity pipeline. His later claim that risk remains partly unknowable also limits any promise that observation completes underwriting. WBUR 2009, 10:49–11:28 and 27:43–28:21
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] Treat venture debt, revenue-based finance, and customer contracts as information-bearing only when repayment and learning milestones are explicit. Analyst-created hypothetical: finance a contracted software vendor through a six-month reporting gate, but do not use debt as a “trial” for a pre-revenue model lab. Current uneven liquidity makes refinancing an assumption to test, not a fallback. Q2 2026 Venture Monitor

2. Match the instrument to cash stage; separate security protection from company risk

• Rule — [researcher inference · strong] Choose debt, preferred, common, or a blend from the company's ability to service fixed claims and the investor's downside objective; never infer that a senior security makes the underlying company safe.
• Evidence status — [investor-stated · strong] for Brooke's 1975 note-plus-common response; [contemporaneous record · strong] for Bancroft's adjacent start-up objection; [researcher inference · moderate] for the generalized rule. 1975 article, printed/PDF p. 6
• Provenance and origin — [investor-stated · strong] In June 1975 Brooke said TA was using notes plus common rather than straight equity amid weak exits. The statement is contemporaneous, but names no company or terms. 1975 article, printed/PDF p. 6
• Mechanism and reasoning — [researcher inference · moderate] A note can create contractual repayment and priority; common preserves upside. The blend works only if cash generation supports debt without starving product, market, or hiring investment. Brooke later made that operating-capacity boundary explicit for buyout leverage. WBUR 2009 exact MP3, 08:19–11:28
• Historical and vehicle context — [contemporaneous record · strong] The 1975 article describes a weak new-issues market, reduced institutional supply, longer holds, and higher follow-on needs. TA exact partnerships and issuers are absent, so this is a market response, not a firm-wide return result. 1975 article, printed/PDF pp. 5–7
• When to use — [researcher inference · moderate] Use debt or blended claims for companies with durable cash flow, measurable coverage, and a repayment path not dependent on a new round. Use equity for uncertainty that cannot bear fixed service. Do not use preference stacks to disguise insolvency or cyclicality.
• How to apply it — [researcher inference · moderate] Model base/downside enterprise value, cash conversion, interest/coverage, maturity, covenants, liquidation stack, dilution, and operating opportunity cost; then test whether the company survives the security under the downside case. Writings modern translation
• Examples
  • 1975 TA financing family — [documented application, investor-stated · strong]. Brooke described notes plus common; issuer, cash stage, and outcomes are unknown. Trade-press scan, printed/PDF p. 6
  • Bank program — researcher-applied illustration. [researcher inference · moderate] Federal-contract collateral could support credit, but the missing default and covenant ledger prevents proof that the security matched every borrower. CHM, printed pp. 17–21
  • Simmons / overleveraged buyouts — [documented boundary, investor-stated · strong]. Brooke condemned serial leverage and admitted debt can prevent products and markets from being funded. WBUR 2009, 06:55–11:28
• Evidence — [researcher inference · strong] Independent 1975 trade press and exact 2009 audio support the stage/capacity boundary; no security-level TA ledger tests performance. 1975 article, printed/PDF p. 6 · WBUR 2009, 06:55–11:28
• Limits, failure modes, and exception ledger — [researcher inference · strong] Priority can redistribute recovery but cannot create demand; debt can make a viable high-growth company fragile; common can overpay for distant upside. Missing TA memos, covenants, defaults, recovery, and cash flows block a universal Brooke policy. Written-record boundary
• Evolution and contradictions — [researcher inference · moderate] The 1975 blend and 2009 anti-overleverage view are compatible when stage differs; without the issuer ledger, they could also reflect a genuine tightening after later credit cycles. Spoken evolution
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] What survives is an integrated downside waterfall; what changes is the wider mix of SAFEs, preferred, venture debt, and structured growth capital amid uneven liquidity. Applicability signals are revenue durability, runway, covenant headroom, refinancing access, and downside recovery; likely misuse is calling debt non-dilutive or a 1x preference “protected” without modeling failure value and foregone growth. Analyst-created hypothetical: compare a revenue-stage software company's preferred round with venture debt under a 12-month liquidity delay, killing the debt option if required product investment breaches covenant headroom. 2026 NVCA Yearbook, financing context

3. Map comparative advantage, then build the relationship funnel around it

• Rule — [investor-stated · strong] Concentrate sourcing where a region, sector, or team has current knowledge and access advantage; build relationships before capital is urgent, then measure whether the channel produces decision-relevant information rather than prestige.
• Evidence status — [investor-stated · strong] for Brooke's comparative-advantage statement and bank funnel; [researcher inference · moderate] for repeatability and performance. Concord transcript, paragraphs beginning “It seemed to me” and “The policy paper”
• Provenance and origin — [investor-stated · moderate] The earliest observed use is Boston technology lending around MIT/Harvard, federal R&D, and regional companies; Brooke later stated the rule directly: finance where the location provides comparative advantage. Concord transcript, paragraphs beginning “The department that I started” and “It seemed to me”
• Mechanism and reasoning — [researcher inference · moderate] Specialized networks lower discovery and diligence costs, improve reference access, and surface change earlier. The advantage decays when the technology base moves, local information becomes commoditized, or the network selects for familiarity rather than truth. CHM, printed pp. 24 and 44–49
• Historical and vehicle context — [investor-stated · moderate] Postwar defense/electronics contracting made Boston technical borrowers unusually visible; later semiconductor migration and minicomputer decline weakened New England knowledge position. Concord transcript, paragraphs beginning “The leaders of industry” and “That advance obsoleted”
• When to use — [researcher inference · moderate] Use when expertise is current, reference networks reach users and operators, and the advantage can be named and tested. Do not use a city, university, family, or firm logo as a proxy for a live edge.
• How to apply it — [researcher inference · moderate] Define the knowledge map; list sources and blind spots; record each source-to-decision fact; compare sourced and non-sourced opportunities; audit conversion and outcomes; retire the thesis when the technical/customer center moves. Sourcing playbook
• Examples
  • Boston bank funnel — [documented application, investor-stated · moderate]. Brooke targeted knowledge-based regional firms and used federal-contract lending to establish relationships. Concord transcript, paragraphs beginning “The department that I started” and “The policy paper that I put”
  • France/Sofinnova — researcher-applied illustration. [researcher inference · moderate] French public institutions and local managers supplied the local leg; Brooke/TA supplied relationships and an international link. The missing Brooke study and disputed dates prevent a cleaner causal claim. Devaux, section 1.2 · CHM, printed pp. 35–36
  • New England platform decline — [documented counterexample, investor-stated · moderate]. Wang and the minicomputer ecosystem show that yesterday’s cluster advantage can become a liability when semiconductor and personal-computer expertise moves. Concord transcript, paragraphs beginning “The leaders of industry” and “That advance obsoleted” · Wang bankruptcy
• Evidence — [researcher inference · strong] Concord and CHM are separate Brooke interview occasions, not independent voices; Devaux supplies external archival scholarship that narrows the France story. No source-to-close or source-to-return ledger survives. Concord transcript, paragraphs beginning “The department that I started,” “It seemed to me,” and “The leaders of industry” · CHM, printed pp. 24 and 35–49 · Devaux
• Limits, failure modes, and exception ledger — [researcher inference · strong] Network homophily, stale expertise, sponsor prestige, geographic essentialism, and selection bias can mimic advantage. Morby's later TA workflow is an explicit non-attribution control: Brooke enabled her hiring, but she built the outbound system. Morby, 15:15–17:28
• Evolution and contradictions — [researcher inference · strong] Brooke expanded the rule from Boston sectors to country affiliates, yet later admitted early-stage technology venture worked abroad mainly in Israel. “Comparative advantage” became stage- and country-specific rather than globally uniform. CHM, printed pp. 44–47
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] A modern sector or geography thesis should predict unique references, data, talent, regulatory interpretation, and decision speed. Likely misuse: branding AI concentration as universal opportunity quality; the 2026 Yearbook's count/value divergence requires a funnel denominator, not a logo narrative. 2026 NVCA Yearbook

4. Test candor and commercializing management, not founder polish

• Rule — [investor-stated · strong] Prefer managers who identify what they do not know, update under challenge, turn technical advantage into a business, and treat evasion as a partnership risk. [researcher inference · moderate] Résumé polish is insufficient unless role-specific evidence supports it. CHM, printed pp. 22–24 and 36–40
• Evidence status — [investor-stated · strong] for the articulated founder-response and management framework; [researcher inference · provisional] for consistent application or predictive efficacy. CHM, printed pp. 22–24 and 36–40
• Provenance and origin — [investor-stated · moderate] Brooke linked the people-reading skill to early private-company work and described testing plan assumptions through founder answers. His later claim that reading people was “really the job” is retrospective, not a contemporaneous rubric. CHM, printed pp. 22–24
• Mechanism and reasoning — [researcher inference · moderate] Candor improves information flow; adaptability lets management revise a flawed plan; commercializing skill connects technology to customers, operations, and finance. The rule fails when interpersonal confidence is mistaken for role competence. A Vision, printed p. 152
• Historical and vehicle context — [researcher inference · moderate] Thin private-company disclosure and long board relationships made management assessment unusually important, but also concentrated judgment in senior sponsors without a visible scoring or bias audit. CHM, printed pp. 22–35
• When to use — [researcher inference · moderate] Use in founder diligence, management succession, and post-investment diagnosis. Do not use social ease, elite credentials, shared background, or agreeable answers as substitutes for work samples and references.
• How to apply it — [researcher inference · moderate] Challenge three load-bearing assumptions; ask what evidence would change each; record unknowns and updates; reference role-specific execution; define the management gap and hire/replace threshold before investing. Unitrode and Wang labs
• Examples
  • Business-plan questioning — [documented application, investor-stated · strong]. Brooke says candor about an unknown improved trust while evasion harmed the prospective partnership. CHM, printed pp. 22–24
  • Unitrode — researcher-applied illustration. [researcher inference · moderate] Brooke and another director reportedly diagnosed a management gap and recruited a leader; the later TI exit does not prove the original diagnosis or sole Brooke causation. CHM, printed pp. 32–33 · TI–Unitrode joint proxy/prospectus, “The Merger — Background” and “Reasons for the Merger,” printed pp. 25–34
  • Wang — [documented boundary, investor-stated · moderate]. A long founder relationship and succession objection did not prevent bankruptcy or platform decline. CHM, printed pp. 27–32 · NYT
• Evidence — [researcher inference · moderate] CHM direct testimony, memoir snippets, and independent company outcomes establish a management/candor theme but not its hit rate. CHM, printed pp. 22–24 and 27–35 · A Vision, printed p. 152 · deal labs
• Limits, failure modes, and exception ledger — [researcher inference · strong] Halo, similarity bias, gender/class filters, hindsight, and board frustration can convert intuition into unjustified certainty. [researcher inference · insufficient public record] This model is below the normal three-distinct-case bar: the generic founder-questioning practice, Unitrode illustration, and Wang boundary are the full surviving set. CHM, the accessible memoir, all 17 named opportunities, company filings, board-record finding aids, and management-change routes were searched; no interview rubric, candidate slate, rejected-manager denominator, or post-hire scorecard surfaced. Deal and retrieval ledger
• Evolution and contradictions — [researcher inference · strong] Brooke called the entrepreneur the hero and investor an enabler, yet described investor-led management changes. The narrow reconciliation is bounded intervention after a defined gap—not a license to override operators whenever performance disappoints. CHM, printed pp. 11–12 and 30–35
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] Candor and commercialization tests survive; AI-assisted storytelling, remote references, and data-governance obligations raise the cost of trusting polish. Applicability signals are explicit assumption updates, customer evidence, reference variance, role simulations, and a written management-gap test; likely misuse is rewarding fluent AI narratives or shared credentials. Analyst-created hypothetical: require an AI-infrastructure founder to revise a sales and data-rights thesis after two contrary customer calls; kill the process if the contradiction is hidden rather than explained. Modern application boundary

5. Pair a local champion with shared standards, explicit authority, and contracted support

• Rule — [researcher inference · strong] Cross-border investing requires a locally accountable decision owner, a common evidence schema, explicit local/central rights, aligned economics, and enforceable support obligations; a brand affiliation or informal promise is not governance.
• Evidence status — [investor-stated · moderate] for local-champion and evolving committee design; [retrospective witness · moderate] for Tang's support failure; [researcher inference · strong] for the combined rule. A Vision, printed pp. 65–71 and 102–113 · Tang, printed pp. 69–72
• Provenance and origin — [investor-stated · moderate] The early international thesis favored locally rooted firms and an accountable champion; later memoir passages describe movement toward paired home-office/local analysis and more central committee authority. A Vision, printed pp. 65–71 and 102–113
• Mechanism and reasoning — [researcher inference · strong] Local teams detect context and relationships; shared standards make cases comparable; explicit votes prevent responsibility gaps; contracted budgets/co-investment make global support credible. Florida/Kenney, printed pp. 119–123 · Hong Kong lab
• Historical and vehicle context — [researcher inference · strong] Sofinnova, Advent UK, country funds, corporate accounts, INF, and pooled regional funds had different capital, owners, mandates, and committees. “Global network” did not denote one legal or economic structure. Vehicle map
• When to use — [researcher inference · moderate] Use for multi-country funds, distributed sector teams, affiliates, and sponsor-supported platforms. Do not launch if the local owner is ceremonial, vetoes conflict, support lacks a budget, or the market lacks viable exits.
• How to apply it — [researcher inference · strong] Write local and central memos; assign sourcing/diligence/board owners; define vetoes, escalation, allocation, reserves, co-investment, response times, information rights, and termination; run a small test before scaling. Japan and Hong Kong labs
• Examples
  • Sofinnova/France — researcher-applied illustration. [researcher inference · moderate] Local public institutions and managers combined with Brooke/TA network links; date, ownership, and decision rights remain disputed. CHM, printed pp. 35–36 · Devaux, section 1.2
  • Japan — [documented counterexample, investor-stated · moderate]. Brooke called the model a failure and attributed the first relationship's end to incompatible diligence/governance; the local countervoice and financial ledger are missing. CHM, printed pp. 46–47 · A Vision, printed pp. 97–98
  • Hong Kong concept — [documented counterexample, retrospective witness · moderate]. Tang says sponsor support and co-investment did not arrive; no public evidence shows deployed capital. Tang, printed pp. 69–72
• Evidence — [researcher inference · strong] The model draws on Brooke’s memoir/direct testimony, Florida/Kenney’s interview synthesis, and independent participant testimony from Tang and Anagnostopoulos. Florida/Kenney, printed pp. 119–123 · Tang, printed pp. 69–72 · Anagnostopoulos, 01:28:00–01:38:00
• Limits, failure modes, and exception ledger — [researcher inference · strong] Centralization can erase local signal; autonomy can fragment evidence and incentives; culture can become a residual explanation; sponsor promises can remain unfunded. Agreements, fund statements, committee minutes, support budgets, and local-manager testimony remain absent for most vehicles after the recorded archive/company/vehicle routes were attempted. Retrieval ledger
• Evolution and contradictions — [researcher inference · strong] The design shifted from affiliate-centered heterogeneity toward integration, but Brooke continued to praise local knowledge. The tension is manageable only through explicit rights, not by choosing “central” or “local” as an ideology. A Vision, printed pp. 102–113
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] Require jurisdiction-specific legal review, data/reporting compatibility, named local ownership, and funded platform commitments. Likely misuse: a global logo plus occasional partner travel. Current U.S. cross-border policy complexity makes the rights matrix transaction-specific. 2026 NVCA Yearbook, policy landscape

6. Make negative information and outside dissent travel

• Rule — [researcher inference · strong] A decision process must preserve bad news, unknowns, and independent challenge from source to final vote; the senior sponsor may respond but may not edit away the dissent.
• Evidence status — [investor-stated · strong] for Brooke's founder-candor and outside-expertise framework; [investor-stated · moderate] for the memoir's committee-information lesson; [researcher inference · provisional] for consistent application. CHM, printed pp. 22–24 and 36–40 · A Vision, printed p. 182
• Provenance and origin — [investor-stated · strong] Brooke's direct framework pairs candid founder answers with current outside technical expertise and dissent; the memoir later condemns withholding information from an IC. CHM, printed pp. 22–24 and 36–40
• Mechanism and reasoning — [researcher inference · moderate] Independent challenge counters stale domain knowledge, sponsor attachment, and consensus pressure; complete packets make decision rights meaningful and preserve error attribution. Federal Express lab
• Historical and vehicle context — [researcher inference · moderate] As TA/Advent added partners, countries, and committees, information became organizational rather than personal. Shared standards could improve transmission, but hierarchy and distance could filter local warnings. A Vision, printed pp. 102–113 and 176–182
• When to use — [researcher inference · strong] Use in every underwriting, follow-on, valuation, management-change, and exit decision. Do not treat a cosmetic risk page or post-terms red team as dissent.
• How to apply it — [researcher inference · strong] Assign an independent reviewer; show missing evidence and source provenance; preserve the dissenter's words; pre-register kill criteria; record the sponsor response, vote, exception, and later postmortem. Decision checklist
• Examples
  • Founder assumption test — [documented application, investor-stated · strong]. Brooke says an honest unknown improved trust while evasion weakened it. CHM, printed pp. 22–24
  • Federal Express follow-on — researcher-applied illustration. [researcher inference · moderate] The story has strong conviction and a named lead, but no dissent/vote/reserve memo; that absence is exactly what the model requires. Lea oral history, Tape 4 00:02:36–00:04:06 · CHM, printed pp. 38–40
  • Japan — [documented counterexample, investor-stated · moderate]. Brooke attributed the first relationship’s end to incompatible diligence and governance. [researcher inference · moderate] The absent local countervoice and decision record make the case a boundary on bad-news transmission, not proof of the witnessed cause. A Vision, printed pp. 97–98
• Evidence — [researcher inference · moderate] Two Brooke direct-voice sources establish the norm; Lea and Japan provide distinct decision boundaries. No complete IC packet or vote log verifies practice. CHM, printed pp. 22–24 and 36–40 · A Vision, printed pp. 97–98 and 176–182 · Lea, Tape 4 00:02:36–00:04:06
• Limits, failure modes, and exception ledger — [researcher inference · strong] Dissent can be low-quality, strategic, or paralyzing; confidentiality can restrict circulation; central templates can erase local nuance. [researcher inference · insufficient public record] This model is below the normal three-distinct-case bar: generic founder questioning, the Federal Express boundary, and Japan are the full surviving set. TA/Advent memos, committee packs, votes, declined-deal logs, Cooksey follow-up material, and the 17-name deal routes were searched; no third named Brooke decision with preserved dissent surfaced. Deal and retrieval ledger
• Evolution and contradictions — [researcher inference · moderate] Brooke promoted outside challenge yet later maintained his Cooksey account after an unrecorded dispute. Preserving the disagreement is evidence of openness; inability to adjudicate it is also a warning about narrative control. CHM follow-up, printed pp. 59–62
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] Independent dissent survives; distributed committees and machine-generated analysis make provenance and decision ownership newly important. Applicability signals are materially different local/domain views, changed forecast inputs, and unresolved model uncertainty; likely misuse is automated risk generation with no evidence threshold. Analyst-created hypothetical: preserve a security specialist's unedited objection to an AI-agent investment, assign an owner to each disputed fact, and defer the vote until the attack-surface test resolves the disagreement. Historical-to-modern boundary

7. Enable through the board without confusing influence with control

• Rule — [researcher inference · strong] Use governance rights for specific, measurable strategy, financing, management, and exit decisions while preserving operator agency; never infer control or causal credit from tenure, advice, or a later successful outcome.
• Evidence status — [investor-stated · strong] for Brooke's enabler framing and claimed board interventions; [documented behavior · strong] for selected director/company events; [researcher inference · moderate] for causal value. CHM, printed pp. 11–12 and 27–35 · Unitrode proxy, “Security Ownership of Certain Beneficial Owners and Management” table
• Provenance and origin — [investor-stated · moderate] Brooke linked his Bessemer move to learning equity ownership and board work, then later articulated the investor-as-enabler boundary. CHM, printed pp. 21–26 and 34–35
• Mechanism and reasoning — [researcher inference · moderate] A board can recruit, structure decisions, finance, and negotiate; it cannot command customers, technology shifts, or founder behavior without actual rights and coalition support. Wang and Unitrode labs
• Historical and vehicle context — [researcher inference · strong] Bank lender, Bessemer representative, TA partner, affiliate sponsor, and Advent director are different roles. A board seat's authority depends on legal owner, security, votes, and fellow directors. People map
• When to use — [researcher inference · strong] Use when a specific intervention fits the board's information, competence, rights, and capacity. Do not intervene merely because the investor is senior, the company is below plan, or hindsight makes another strategy obvious.
• How to apply it — [researcher inference · strong] Name the problem, decision owner, board right, evidence, alternatives, vote, implementation owner, milestone, cost, and exit/stop; separate advice from execution and outcome. Governance checklist
• Examples
  • Unitrode — [documented application, investor-stated · moderate]. Brooke and another director reportedly recruited new management; the co-director and board vote are missing, and the later TI deal is not causal proof. CHM, printed pp. 32–33 · TI–Unitrode joint proxy/prospectus, “The Merger — Background” and “Reasons for the Merger,” printed pp. 25–34
  • Adams Russell — [documented application, investor-stated · provisional]. Brooke says he proposed separating electronics and cable; contemporaneous reports show company events, not Brooke's implementation or proceeds. CHM, printed p. 32 · 1986 report, printed p. 57
  • Wang — [documented counterexample, investor-stated · moderate]. Long service and succession advocacy did not prevent platform decline and bankruptcy; formal authority and feasible alternatives are unknown. CHM, printed pp. 27–32 · NYT
• Evidence — [researcher inference · strong] CHM direct testimony is triangulated with Unitrode filings, Adams Russell transaction reporting, and Wang bankruptcy reporting. The intervention-to-outcome bridge remains incomplete. CHM, printed pp. 27–35 · Unitrode proxy, “Security Ownership of Certain Beneficial Owners and Management” table · Adams Russell report, 30 June 1986, printed p. 57 · Wang bankruptcy report, 19 August 1992
• Limits, failure modes, and exception ledger — [researcher inference · strong] Board theater, sponsor overreach, management scapegoating, unclear votes, insufficient domain competence, and outcome attribution are core risks. Missing minutes, cap tables, votes, intervention budgets, and counterfactual plans block efficacy claims after the board/archive routes were attempted. Retrieval ledger
• Evolution and contradictions — [researcher inference · strong] Brooke's “enabler” humility and active interventions are compatible only if help remains bounded and attributed. Aspen's “we” narrative and Brown's filed board role show why institutional language is not individual causation. Florida/Kenney, printed pp. 122–123 · Aspen 424B4, “Management”
• Modern VC translation (as of 2026-08-18) — [researcher inference · strong] Bounded board help survives; larger platform teams and denser service claims make owner, baseline, cost, right, and counterfactual more important. Applicability signals are a named company need, board authority, measurable milestone, and management consent; likely misuse is marketing access or recruiting as value creation without causal evidence. Analyst-created hypothetical: pre-register a pricing intervention for a vertical-software company and reject sponsor credit unless the board-approved owner, cohort baseline, cost, and six-month milestone are recorded. Governance evidence

8. Pre-register operating value; do not substitute leverage or narrative

• Rule — [researcher inference · strong] Underwrite product, market, management, working-capital, and strategic actions before entry; treat leverage, entry price, fees, market beta, and exit multiple as separate return components.
• Evidence status — [investor-stated · strong] for Brooke's operating-value and leverage boundaries; [researcher inference · insufficient public record] for portfolio-wide efficacy. WBUR 2007, Brooke 09:36–15:53 · WBUR 2009, Brooke 02:29–11:28
• Provenance and origin — [investor-stated · provisional] Brooke’s 1993 Aspen account named acquisitions, introductions, hiring, and downsizing; the 2007/09 audio later stated the broader operating-value-versus-leverage rule. Florida/Kenney, printed pp. 122–123 · WBUR 2007, 09:36–15:53
• Mechanism and reasoning — [researcher inference · strong] Only an action with a baseline, responsible owner, cost, timing, milestone, and counterfactual can support causal value; leverage and market reopening may amplify return without improving the company. Blaydon independently decomposed buying price, debt, and cash-flow improvement. WBUR 2007, 27:43–29:59
• Historical and vehicle context — [investor-stated · strong] Brooke spoke after a cheap-credit buyout boom and before/after the financial crisis; he distinguished venture equity, expansion/private equity, and leveraged buyouts. These were different capital structures, not one industry mechanism. WBUR 2009, 02:29–11:28
• When to use — [researcher inference · strong] Use in every buyout, growth, board-value, platform, and international-expansion case. Do not count multiple expansion, refinancing, layoffs, or a company acquisition as value creation without the operating bridge.
• How to apply it — [researcher inference · strong] Decompose entry price, financing, fees, operations, organic/acquired growth, employment/product investment, market movement, and exit; pre-register each initiative; compare plan with realized KPI and cost. Aspen lab
• Examples
  • Aspen — [documented application claim, investor-stated · provisional]. Brooke attributed cross-border actions to Advent, while Brown's board role and management execution constrain credit; no causal return bridge exists. Florida/Kenney, printed pp. 122–123 · Aspen filing, “Management”
  • Terapia — researcher-applied successor illustration. [researcher inference · moderate] IFC describes management, environmental, and governance work in a 2003 successor-era deal; it tests the mechanism but is excluded from Brooke performance. IFC case, printed/PDF p. 32 onward
  • Simmons / caller companies — [documented counterexamples, contemporaneous record · moderate]. Debt and fees depleted companies in the panel accounts; Brooke called such conduct indefensible. Unnamed cases lack audit-level facts. WBUR 2009, 06:55–09:22 and 33:08–38:06
• Evidence — [researcher inference · strong] Two exact-audio programs, Florida/Kenney, an SEC filing, and IFC establish mechanisms and counterarguments; no portfolio decomposition establishes prevalence or return attribution. WBUR 2007, 09:36–15:53, 18:56–24:10, and 33:19–38:32 · WBUR 2009, 03:58–11:28, 23:19–33:01, and 35:00–38:06 · IFC Terapia, printed/PDF p. 32 onward
• Limits, failure modes, and exception ledger — [researcher inference · strong] Sponsor self-report, selection of winners, undefined “value,” hidden acquisition cost, employment displacement, underinvestment, and market beta can launder causation. Missing Advent initiative ledgers and portfolio cash-flow bridges prevent validation after interview, filing, public-LP, and board-record routes were attempted. Retrieval ledger
• Evolution and contradictions — [researcher inference · moderate] The 1975 use of debt and later anti-overleverage stance can coexist when cash stage differs. Brooke also defended a majority of managers without data while admitting destructive cases, leaving prevalence unresolved. WBUR 2007, 11:30–15:53 and 33:39–34:20
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] Initiative-level attribution survives; AI automation, add-on acquisitions, and constrained refinancing add labor-transfer, integration, and capital-structure assumptions. Applicability signals are a pre-action KPI baseline, named owner, cost, product/customer effects, and a separable exit case; likely misuse is calling cuts or acquisitions durable value without those bridges. Analyst-created hypothetical: underwrite an AI support rollout only after separating gross-margin improvement from churn, implementation cost, displaced labor, leverage, and exit-multiple change. Operating-value boundary

9. Separate mandates and risk budgets; treat exit liquidity as a portfolio input

• Rule — [researcher inference · strong] Keep early venture, growth, buyout, corporate strategic, geography, and learning capital in explicit sleeves with their own loss limits, reserves, manager capacity, and exit assumptions; test results with and without the dominant winner.
• Evidence status — [investor-stated · moderate] for diversified/stage-separated practice and INF stage-drift postmortem; [investor-stated · strong] for exit drought; [researcher inference · strong] for the combined portfolio rule. A Vision, printed pp. 26 and 114–117 · Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s”
• Provenance and origin — [investor-stated · moderate] The memoir attributes a diversified, revenue-oriented TA posture to the early platform; Brooke’s 1994 interview separately makes exit windows part of realized performance. A Vision, printed p. 26 · Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s”
• Mechanism and reasoning — [researcher inference · strong] Separating sleeves prevents a mission or hot stage from importing unpriced risk; reserves and manager maturity become constraints; exit assumptions connect company value to realizable cash. INF lab
• Historical and vehicle context — [documented behavior · strong] A September 1996 IFC retrospective analysis reported poor aggregate early emerging-market fund outcomes and linked small funds, inexperienced managers, weak exits, and governance to problems. [researcher inference · strong] The later analysis does not prove INF belonged to its sample or describe contemporaneous INF underwriting. IFC Note 90, printed pp. 2–4, September 1996
• When to use — [researcher inference · strong] Use whenever one fund mixes stages, countries, experimental mandates, or strategic capital. Do not aggregate unlike vehicles or infer liquidity from a company's operating progress.
• How to apply it — [researcher inference · strong] Define sleeve capital/count, expected loss, reserves, manager capacity, ownership, exit path, and stop gate; report gross/net and realized/unrealized separately; calculate ex-top-winner sensitivity when denominators permit. Portfolio forensics
• Examples
  • 1970s TA portfolio — [documented boundary, investor-stated · strong]. Brooke remembered ten to eleven years without substantial realization before 1979–1981 reopening. Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s”
  • INF — [documented counterexample, investor-stated · moderate]. Memoir says a global pool took too much early-stage risk and produced weaker results; no audited cash flows survive. A Vision, printed p. 114
  • ESSF/GPE II — [institutional controls, investor-stated · provisional]. Manager-reported net results are positive but whole-fund, team, and unaudited; no company or Brooke bridge exists. Advent presentation, PDF p. 17
• Evidence — [researcher inference · strong] Brooke’s two direct corpora, IFC, and manager-reported fund controls provide independent boundaries; the true Brooke deal/portfolio denominator is not computable. Concord transcript, paragraphs beginning “Going back a little bit” and “It wasn’t until the late ‘70s” · IFC Note 90, printed pp. 2–4, September 1996 · denominator
• Limits, failure modes, and exception ledger — [researcher inference · strong] Sleeve labels can conceal correlated risks; reported IRR can obscure DPI, fees, and marks; reserves can become escalation. Twenty-five vehicle families yield zero Brooke-personal returns, and 17 named opportunities yield zero complete reconstructions. Vehicle and deal denominators
• Evolution and contradictions — [researcher inference · strong] Brooke's diversification/randomness language sits beside Federal Express follow-on faith. Conviction is compatible only when a named sleeve, reserve owner, new evidence, and loss budget permit the exception. CHM, printed pp. 11, 22–24, and 38–40
• Modern VC translation (as of 2026-08-18) — [researcher inference · strong] Report count and capital concentration separately; model AI/mega-round exposure, reserves, and exits as explicit overlays. The 2026 Yearbook's 3.2%-of-count/67%-of-value mega-deal split illustrates why a dollar boom is not broad-based. 2026 NVCA Yearbook

10. Develop whole-job investors and make team succession visible

• Rule — [researcher inference · moderate] Develop investors across sourcing, diligence, portfolio work, exits, and fundraising; make role progression, shared economics, failure rehabilitation, confidence checks, and succession explicit so the institution can outlive a founder.
• Evidence status — [investor-stated · strong] for Brooke's coaching and whole-job claims; [retrospective witness · strong] for Morby's role boundary; [researcher inference · provisional] for institutional efficacy. CHM, printed pp. 42–43 and 53–54 · Morby, 08:46–17:28
• Provenance and origin — [investor-stated · moderate] Brooke later connected team chemistry to school sports and described coaching investors through both failure and early success. The causal link from sports is unproved; the management rule is direct. CHM, printed pp. 5–6, 42–43, and 53–54
• Mechanism and reasoning — [researcher inference · moderate] Rotations build judgment and attribution; shared upside improves retention; visible succession prevents founder bottlenecks; coaching prevents one loss or win from dominating behavior. A Vision, printed pp. 176–193
• Historical and vehicle context — [retrospective witness · moderate] TA’s later domestic system developed under Landry and partners; Advent had to rebuild a U.S. team; compensation and ownership schedules remain private. French, printed pp. 59–61 · CHM, printed pp. 40–43
• When to use — [researcher inference · strong] Use before scaling fund size, sectors, boards, or geographies. Do not rely on culture slogans when partner capacity, ownership, advancement, or key-person transitions are opaque.
• How to apply it — [researcher inference · moderate] Record roles by decision stage; rotate whole-job work; cap board load; pair success/failure reviews with specific development; publish advancement criteria; test succession before key-person departure. Team map
• Examples
  • Morby at TA — [documented application, retrospective witness · strong]. Brooke hired and enabled; Morby created/taught sourcing; Landry reviewed investments. The explicit division is the model's attribution requirement. Morby, 08:46–17:28 and 32:59–35:18
  • TA teamwork failure — [documented counterexample, investor-stated · moderate]. Brooke later said failure to build teamwork contributed to good people leaving; employment/economics records are missing. A Vision, printed p. 193
  • Advent succession — [documented boundary, contemporaneous record · strong]. Brown became CEO in January 1996 by filing, while Brooke remembered 1997; role transitions need legal and operating dates. Aspen 424B4, “Management” · CHM, printed p. 58
• Evidence — [researcher inference · strong] CHM, memoir, French, Morby, and a contemporaneous SEC filing supply independent role/succession evidence; compensation, promotion, retention, and performance denominators remain absent. French, printed pp. 59–61 · Morby, 08:46–17:28 and 32:59–35:18 · Aspen filing, “Management”
• Limits, failure modes, and exception ledger — [researcher inference · strong] Rotations can produce generalists without edge; shared economics can entrench incumbents; mentoring can become patronage; succession titles can precede real authority. Partnership agreements, carry/vesting, departure reasons, and workload histories remain missing after employment, oral-history, filing, and agreement routes were attempted. Retrieval ledger
• Evolution and contradictions — [researcher inference · strong] Brooke advocated teamwork while admitting failure at TA; founding Advent required another team rebuild. The contradiction supports process humility, not proof that later Advent solved every incentive problem. CHM, printed pp. 40–43 · A Vision, printed p. 193
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] Audit deals/partner, active boards, sector competence, source credit, carry/vesting, and next-generation votes before fund growth. Likely misuse: assigning all wins to a brand-name founder while describing losses as team decisions. ILPA Principles 3.0, alignment and key-person guidance, printed pp. 10–22 and glossary p. 41

11. Match vehicle, incentives, authority, and duration to the strategy

• Rule — [researcher inference · strong] Select the capital form, fund life, mandate, governance, economics, geographic rights, regulation, liquidity, and succession terms that permit the strategy; change the vehicle when those conditions no longer fit.
• Evidence status — [documented behavior · strong] for Brooke's passage through distinct institutions and filed entities; [researcher inference · moderate] for why each transition occurred and affected outcomes. CHM, printed pp. 16–50 · federal notices
• Provenance and origin — [researcher inference · moderate] The earliest observed transition is bank debt to Bessemer family equity in 1961; later Tucker/TA pooling and Advent's cross-border fork expanded capital and authority. Brooke narrated these moves retrospectively. CHM, printed pp. 21–27 and 40–43
• Mechanism and reasoning — [researcher inference · strong] Vehicle design determines who supplies capital, time horizon, approval, geography, reserves, information, compensation, liquidity, and key-person continuity; a philosophical strategy cannot override incompatible legal/economic incentives. Vehicle analysis
• Historical and vehicle context — [contemporaneous record · strong] Devonshire and Advent IV notices show different capital owners and partners. 1975 notice, printed p. 42805 · 1981 notice, printed p. 52267 [retrospective witness · moderate] Anagnostopoulos adds participant-reported Advent UK/global forms; together the records show heterogeneity, not a consolidated Brooke vehicle. Anagnostopoulos, 01:28:00–01:38:00
• When to use — [researcher inference · strong] Use before every new strategy, geography, account, annex, SPV, continuation, or successor fund. Do not infer common economics or authority from a shared brand.
• How to apply it — [researcher inference · strong] Build a term matrix for capital owner, legal form, duration, fees/carry, GP commitment, decisions/vetoes, allocation, reserves, liquidity, reporting, geography, regulation, key person, succession, and termination; test each against the strategy. Vehicle ledger
• Examples
  • Bank to Bessemer — [documented behavior, investor-stated · moderate]. Role shifted from creditor inside bank approval to employee/director inside family capital; personal ownership remains unknown. CHM, printed pp. 16–26
  • TA to Advent — [documented transition, retrospective witness · moderate]. Partners rejected the global path; relationships and people split; Landry continued TA. Agreements and asset allocations are absent. CHM, printed pp. 40–43 · French, printed pp. 59–61
  • Hong Kong concept — [documented counterexample, retrospective witness · moderate]. Tang says he withdrew after expected sponsor support and co-investment did not arrive. No deployment, agreement, committee record, or authority document survives, so the witness account does not establish a single cause. Tang, printed pp. 69–72
• Evidence — [researcher inference · strong] Oral histories, federal notices, UK filings, SEC filings, and participant testimony establish vehicle heterogeneity; 25 named families were searched, four permit narrow hard observations, and zero supports a Brooke-personal return. 1975 notice, printed p. 42805 · 1981 notice, printed p. 52267 · vehicle denominator
• Limits, failure modes, and exception ledger — [researcher inference · strong] Brand continuity, affiliate labels, AUM, or founder title can hide different owners and rights. Partnership agreements, allocation policies, fee/carry schedules, LP cash flows, key-person terms, and formation files remain missing after federal, UK-company, SEC, participant, archive, and public-LP routes were attempted. Vehicle retrieval ledger
• Evolution and contradictions — [researcher inference · strong] Brooke's institution-building required delegation, yet voting control and centralization appear in later accounts. Local entrepreneurship and founder continuity remained in tension with scalable governance. A Vision, printed pp. 54, 79, and 102–113
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] Treat SPVs, opportunity funds, continuation vehicles, and regional affiliates as separate incentive systems. Likely misuse: assigning successor-fund or co-investment results to a founder. Use ILPA's alignment, governance, reporting, and key-person questions as diligence prompts, not laws or market averages. ILPA Principles 3.0, printed pp. 10–22 and glossary p. 41

12. Ledger person, vehicle, instrument, and decision separately

• Rule — [researcher inference · strong] For every opportunity, record the legal investor, capital owner, source, sponsor, diligence, approval, negotiation, board work, follow-on, exit, security, ownership, cash flow, and later narrator separately; never turn “we,” a firm logo, or founder status into personal attribution.
• Evidence status — [researcher inference · strong] The rule is reconstructed from repeated attribution conflicts across direct testimony, participant accounts, and filings; it is not a Brooke-stated maxim. People map · ledger
• Provenance and origin — [researcher inference · strong] The earliest record-wide need appears when Bank capital, Bessemer work, and later board roles are collapsed; FedEx, Biogen, Morby, Advent UK, and Aspen make the rule unavoidable. CHM, printed pp. 27–43
• Mechanism and reasoning — [researcher inference · strong] Role separation prevents hero stories, clarifies decision rights, connects economics to the correct entity, and makes learning assignable. Without it, neither skill nor accountability can be tested.
• Historical and vehicle context — [contemporaneous record · strong] Federal notices, proxies, prospectuses, and public-LP materials expose distinct legal owners and entities. Unitrode proxy, “Security Ownership of Certain Beneficial Owners and Management” table · Advent presentation, PDF p. 17 [retrospective witness · moderate] Oral histories add actor claims that require separate attribution; disciplined “unknown” remains essential. People map
• When to use — [researcher inference · strong] Use for every deal, fund, board claim, return, pass, and successor narrative. Do not use a company outcome to infer investor proceeds or a board seat to infer sourcing.
• How to apply it — [researcher inference · strong] Create dated rows for entity, check, security, stake, approvers/objectors, sponsor, board representative, reserves, follow-ons, transfers, distributions, and narrator; reconcile aliases and successor vehicles before calculating. Denominator method
• Examples
  • Federal Express — [documented correction, retrospective witness · strong]. Lea was principal lead; another TA colleague increased the final-round position; Brooke does not claim either. Lea oral history, Tape 4 00:02:36–00:04:06 · CHM, printed pp. 38–40
  • Biogen. [contemporaneous record · strong] TA participated and Landry had a filed board role. [investor-stated · provisional] reported: ~50x (Boston Globe obituary, 2020), unranked and unaudited. [researcher inference · strong] No Brooke decision or cash-flow proof survives. BHR study, printed p. 203 n. 71/PDF p. 18 · Landry SEC biography, “Kevin G. Landry” paragraph
  • Aspen. [investor-stated · provisional] Brooke said “we.” [retrospective witness · moderate] A founder reports Advent financing. [contemporaneous record · strong] A filing names Brown as director. [researcher inference · strong] No personal Brooke return or causal bridge survives. Florida/Kenney, printed pp. 122–123 · Aspen filing, “Management”
• Evidence — [researcher inference · strong] Lea, Morby, Anagnostopoulos, French, company filings, and fund records are independent controls against founder-wide attribution. Morby, 15:15–17:28 · Anagnostopoulos, 01:28:00–01:38:00 · French, printed pp. 59–61
• Limits, failure modes, and exception ledger — [researcher inference · strong] Records may omit informal influence; legal ownership may not equal decision work; witnesses may self-credit. The answer is parallel fields and uncertainty, not choosing one narrator. Seventeen named opportunities and 25 vehicle families still yield zero personal return reconstructions. Evidence boundary
• Evolution and contradictions — [researcher inference · strong] Brooke often credited others and called investors enablers, yet retrospective institutional narratives still use “we” and founder-centered shorthand. The rule preserves his humility more faithfully than assigning all branded outcomes to him. CHM, printed pp. 11–12 and 38–43
• Modern VC translation (as of 2026-08-18) — [researcher inference · strong] The attribution ledger survives; proliferating SPVs, opportunity funds, continuations, personal positions, and operating teams create more allocation and conflict boundaries. Applicability signals are overlapping vehicles, shared deals, partner transitions, co-investment, or brand-level claims; likely misuse is treating a portfolio page or partner bio as cash-flow proof. Analyst-created hypothetical: for a continuation transaction, record the selling and buying vehicles, decision voters, personal interests, operating work, price process, and distributions before assigning the result to any partner. 2026 NVCA Yearbook, vehicle context

13. Predefine defeat, follow-on gates, and the learning budget

• Rule — [researcher inference · strong] Before entry, define evidence that kills or narrows the thesis, who can stop follow-ons, how much capital may be spent learning, and which postmortem change must follow failure.
• Evidence status — [investor-stated · moderate] for admitting defeat, Japan, and INF postmortems; [retrospective witness · moderate] for Hong Kong; [researcher inference · moderate] for a consistent historical stop process. A Vision, printed pp. 97–127 and 177 · Tang, printed pp. 69–72
• Provenance and origin — [investor-stated · moderate] The developed memoir says managers should admit defeat rather than hide it and offers Japan/INF/team failures; no earlier formal stop rubric surfaced. A Vision, printed p. 177
• Mechanism and reasoning — [researcher inference · strong] Precommitment limits escalation, distinguishes learning from rationalization, preserves reserves, and forces evidence to update stage/geography/team design. Failure matrix
• Historical and vehicle context — [researcher inference · strong] Cross-border experiments combined uncertain managers, thin exits, currency risk, and mission pressure; without sleeve limits, a learning rationale could consume core return capital. IFC Note 90, printed pp. 2–4, September 1996
• When to use — [researcher inference · strong] Use for every initial check, reserve, bridge, rescue, new geography, and platform commitment. Do not let category prestige, sunk cost, or sponsor reputation replace changed evidence.
• How to apply it — [researcher inference · strong] List thesis falsifiers, milestone date, evidence owner, reserve ceiling, stop/restructure authority, recovery options, and postmortem output; every follow-on memo must compare new evidence with the original gate. Failure decision labs
• Examples
  • Japan — [documented counterexample, investor-stated · moderate]. The first relationship ended and Brooke called the model a failure; invested capital and stop decision are unknown. CHM, printed pp. 46–47 · A Vision, printed pp. 97–98
  • INF — [documented counterexample, investor-stated · moderate]. Excess early-stage risk reportedly weakened results; no stage limit or reserve record survives. A Vision, printed p. 114
  • Hong Kong concept — [documented boundary, retrospective witness · moderate]. Tang left after expected support failed; no deployed capital is shown, making this an organizational stop rather than a realized loss. Tang, printed pp. 69–72
• Evidence — [researcher inference · strong] Memoir, CHM, Tang, and IFC provide multiple sources and distinct failures; no complete write-off, reserve, bridge, or pass ledger establishes routine application. CHM, printed pp. 38–47 · Tang, printed pp. 69–72 · IFC Note 90, printed pp. 2–4, September 1996
• Limits, failure modes, and exception ledger — [researcher inference · strong] Premature stopping can destroy option value; milestones can be gamed; learning can be unfalsifiable. Vehicle cash flows, IC minutes, write-down dates, rescue decisions, and postmortem changes remain missing after memoir, archive, fund, participant, and filing routes were attempted, keeping the model moderate. Retrieval ledger
• Evolution and contradictions — [researcher inference · strong] Brooke praised Federal Express conviction while later emphasizing defeat. The coherent boundary is evidence-backed persistence inside a predeclared reserve—not faith after the learning budget is exhausted. CHM, printed pp. 38–40 · A Vision, printed p. 177
• Modern VC translation (as of 2026-08-18) — [researcher inference · moderate] Require a capped bridge budget, milestone evidence, ownership gained per dollar, and explicit write-down/stop authority. Likely misuse: calling repeated insider bridges “learning” while liquidity and concentration worsen. Q2 2026 Venture Monitor

Evidence gaps and candidate models

The thirteen models reconcile the distinct rules supported across the note files, written corpus, spoken corpus, deal forensics, and philosophy index. [researcher inference · strong] Credit-as-information is separate from instrument fit because the first concerns learning and the second fixed-claim survivability; local champion, shared standards, and sponsor support remain one governance system because each is necessary to make distributed authority executable.

Plausible candidates excluded from the model inventory:

[researcher inference · strong] The public record is exhausted enough to keep these as exclusions rather than unfinished models: the highest-value remaining evidence is archival or owner-controlled. Recovery of the 1971/1973 bodies, bank files, TA/Advent agreements and IC records, complete A Vision text, both Done Deals chapters, board records, passes, write-offs, and cash flows could add, merge, split, or materially downgrade models. Source-map coverage assessment

What the evidence supports

[researcher inference · strong] The public record supports an institution-builder thesis more strongly than a personal-alpha thesis. Brooke repeatedly moved among a bank credit program, an existing family office, TA partnerships, and a cross-border Advent platform, while the surviving evidence separates those vehicles and names the partners, local managers, executives, and boards who owned material work. The same record does not provide a complete Brooke-personal cash-flow series, hit rate, or attribution bridge. Profile edge and decomposition · vehicle and deal denominators

[researcher inference · strong] The source base is unusually explicit about its limits: 88 mapped source rows comprise 41 ★★★, 43 ★★, and 4 ★ items; 84 priority rows reconcile to phase ownership. The WRITINGS record contains ten note families but only two full target bodies and one partial body; the TALKS record contains eleven priority families, including six full oral-history transcripts, three exact-file audio appearances, one exact-file captioned video, and one forum family with no authenticated stage body. These are coverage denominators, not independent-source counts or proof that inaccessible bodies say what their titles imply. Source-map coverage · written-record denominator · spoken-corpus coverage

The best-supported decision lessons are bounded rather than heroic:

• Stage and instrument — [investor-stated · strong; researcher inference · moderate]. Brooke's 1975 note-plus-common response is real, while Bancroft's adjacent objection makes cash stage a necessary boundary; no issuer-level outcome ledger proves a universal policy. 1975 article, printed/PDF p. 6
• Local information plus explicit authority — [investor-stated · moderate; researcher inference · strong]. Brooke's local/global hypothesis survives as a governance question, not a return theorem; Japan, INF, ChinaVest, and Tang's Hong Kong account are the controls. CHM oral history, printed pp. 44–47 · Tang oral history, printed pp. 69–72
• Team attribution — [researcher inference · strong]. Lea, Landry, Morby, Anagnostopoulos, Cooksey, Brown, local managers, founders, and boards cannot be collapsed into Brooke or an Advent logo. People map
• Operating value — [investor-stated · strong; researcher inference · insufficient public record]. Brooke argued for operating improvement and against leverage-only narratives, but his own admissions and WBUR countervoices require a deal-level bridge from baseline to action to outcome. WBUR 2007 exact MP3, 09:36–15:53 and 16:34–36:35 · WBUR 2009 exact MP3, 03:58–11:28 and 20:22–38:06

What could falsify the thesis

The favorable thesis predicts that unseen ordinary-case files will repeatedly show Brooke-specific local information changing approvals or terms, stage-matched capital, named dissent, explicit local/central authority, operating milestones with owners, and better ordinary outcomes when those mechanisms were present. [researcher inference · moderate] Profile predicted evidence

It weakens or fails if original records instead show generic selection, partner-led execution with little Brooke input, no decision effect from local information, interventions that did not change outcomes, weak results excluding the top winner, or economics driven mainly by privileged access, leverage, the 1979–1981 reopening, and successor teams. [researcher inference · strong] The current 17-company/opportunity set has zero complete Brooke-attributable entry-to-exit reconstructions, so the public record cannot decide that contest. Denominator reconciliation · strongest skeptical case

The thesis is also vulnerable to three non-return tests:

1. Causal chronology: Bank, Army, Marshall Plan, or Bessemer exposure must not be treated as proof of a later doctrine without a dated bridge. Profile formative boundary
2. Portable governance: a local champion must have documented economics, authority, escalation, and support; a flag on a map is not a replicated capability. Model 5
3. Institutional persistence: successor scale or later-fund results must not be assigned backward to Brooke without role, period, legal-entity, and cash-flow continuity. Model 12

Attribution and denominator audit

The five material attribution checks required by REVIEW pass with bounded conclusions: Federal Express belongs first to Charlie Lea and the later TA follow-on decision maker; later TA sourcing belongs to Morby/Landry/team; Biogen has TA and Landry controls but no Brooke underwriting/cash-flow proof; Aspen includes Brown, management, and an Advent vehicle without a personal causal bridge; Sofinnova and Advent UK require local/institutional co-credit. Lea oral history, Tape 4 00:02:36–00:04:06 · Morby exact MP3, 15:15–17:28 · Landry filed biography, “Kevin G. Landry” paragraph · Aspen filing, “Management” · Anagnostopoulos, 01:28:00–01:38:00

Transfer limits for modern VCs

[researcher inference · strong] The package transfers procedures, not historical performance claims. A current analyst can use the stage/instrument test, attribution ledger, dissent channel, local/central rights map, operating-value bridge, vehicle fit test, and predeclared stop gates. The analyst cannot import Brooke-era bank collateral, thin 1970s exit markets, family-office duration, SBIC structures, country-affiliate authority, or a manager's later fund economics as present-day base rates. Philosophy modern translation · mental-model field guide

Current-market premises are dated and used narrowly. The Q2 2026 Venture Monitor and 2026 NVCA Yearbook establish concentration and uneven liquidity observations, not the truth or efficacy of a Brooke rule. ILPA Principles 3.0 supplies best-practice comparison dimensions, not law or a measured market average. [researcher inference · strong] Writings translation, as of 2026-08-02 · dimension-matched benchmark

Likely modern misuses are: calling debt non-dilutive without downside coverage; treating a famous local partner as authority; converting a board seat into causation; pooling unlike mandates; reading IRR without DPI, marks, fees, and period; assigning a firm result to one person; or laundering mission into return. The reusable checklists explicitly guard those errors. Analyst checklists

Unresolved evidence

The remaining evidence gaps are material, named, and mostly archival or owner-controlled after the documented public-route sweep:

• Bank policy, credit-committee, borrower, default, recovery, and later-equity records;
• Bessemer/Tucker holdings and TA/Advent agreements, IC votes, allocations, LP reports, passes, write-offs, reserves, and cash flows;
• continuous access to A Vision, the 1971 and 1973 article bodies, the CODIMEC proceedings, both Done Deals chapters, and any separate French/Arthur D. Little artifact;
• Wang, Unitrode, Biogen, Aspen, Japan, INF, ChinaVest, and Hong Kong decision/ownership/cash-flow bridges; and
• independent countervoices for disputed local-manager and departure accounts.

Written retrieval agenda · deal retrieval ledger · source-map archive routes

[researcher inference · strong] These gaps do not justify a filler continuation: the package names the legitimate routes tried, the access barrier, and the claim each missing record could change. They do prohibit a computed personal record, audited causal attribution, and a universal Brooke doctrine. Written routes · deal routes · source-map routes

Central-claim citation audit trail

Ordinary-claim citation sample

Sampling rule: within each of the seven pre-review root documents, enumerate externally checkable, claim-bearing blocks with a claim-local link in source order after excluding headings, navigation-only lines, and central claims already audited above; inspect the first block at or after 25% and the first block at or after 75% of that document's eligible sequence. This deterministic spread yields two claims per document (14 total).

Twenty-point acceptance evidence

Source-ranking protocol

Priority is retrieval order, not an endorsement: ★★★ must process; ★★ should process; ★ process if time. Reliability A denotes direct testimony or an authoritative record, B an institutional or quality secondary account, and C a discovery lead, interested paraphrase, or unprocessed restricted source. Access and locator notes record what was actually inspected during SCOUT.

Writings

• ★★★ | A Vision for Venture Capital: Realizing the Promise of Global Venture Capital and Private Equity | https://www.econbiz.de/10009139826 | 2009 | coauthored book catalog and partial preview; reliability C until full text | ISBN 9781584657996, xviii plus 275 pp.; Brooke and Daniel Penrice are credited, only partial text was accessible, and later witness challenges require page-level checking.
• ★★★ | How a Private Venture Capital Firm Operates | https://www.jstor.org/stable/40685219 | 1971-01 | The Business Lawyer 26(3), pp. 735–738; reliability A for publisher metadata | JSTOR's citation record credits Peter A. Brooke alone and supplies DOI 10.2307/40685219; the body remains subscription-gated, so the title cannot substitute for its argument.
• ★★★ | Should Pension Funds Participate in Venture Capital? | https://publications.gc.ca/collections/collection_2023/isde-ised/ST31-4-1975-eng.pdf | 1973-09/10 | original article reported in a 1975 Government of Canada bibliography; reliability A for the citation | Printed p. 75/PDF p. 85 credits Peter A. Brooke and Pensions, New York; the serial identity, pagination, and body remain unresolved and must not be normalized to Pensions & Investments.
• ★★ | Histórico da TA Associates | https://repositorio.fgv.br/bitstreams/6e9e7b7b-e9e6-4f55-80b8-06fc20fd5cc1/download | 1981 | later FGV-hosted academic bibliography; reliability C for original contribution | Credits “P. Brooke” in CODIMEC's 1° Seminário Internacional sobre Venture Capital white paper, but no original proceedings, page range, or evidence distinguishing authored paper, translation, slides, or edited speech surfaced; upgrade only if the original confirms a Brooke paper.

Talks and video

• ★★★ | Peter Brooke: Private Equity Now | https://www.wbur.org/onpoint/2009/11/24/peter-brooke | 2009-11-24 | radio interview/panel and downloadable audio; reliability A for participant speech | The live Megaphone MP3 is 22,467,013 bytes, runs 46:41.946, and has SHA-256 95e7f28d9abff91c1a29f8795f5cad37a735b48f4742166158786b28d04fe34e; WBUR displays 24:22, so every timestamp binds to the exact MP3 rather than the unresolved page duration.
• ★★★ | Big deals in private equity | https://www.wbur.org/onpoint/2007/06/25/big-deals-in-private-equity | 2007-06-25 | radio panel and audio; reliability A for participant speech | The live Megaphone MP3 is 21,037,435 bytes, runs 43:43.216, and has SHA-256 4399badb5a63b95804bdd0823feabcd0d8507b7be0ad0796d786882eae6b371f; Brooke appears with labor, policy, academic, reporter, and caller countervoices.
• ★★ | Peter Brooke and Terry McGuire discuss a global vision for venture capital and private equity | https://www.wilmerhale.com/en/insights/events/xconomy-peter-brooke-and-terry-mcguire-discuss-a-global-vision-for-venture-capital-and-private-equity-october-21-2009 | 2009-10-21 | WilmerHale sponsor record plus recovered contemporaneous Xconomy reporting; reliability B | Xconomy presented the forum; WilmerHale sponsored it and provided the venue. A complete 2009-09-24 Xconomy announcement contains a separate 2009-09-23 reporter interview with Brooke; the complete 2009-10-20 preview, complete 2009-10-29 post-event gallery, and a named attendee recap also survive. The September interview is not October 21 stage testimony or a separate priority-appearance family; authenticated forum audio/transcript remains unavailable.
• ★★ | 40 years of growth, 40 years of partnership | https://www.linkedin.com/posts/advent-international_40-years-of-growth-40-years-of-partnership-activity-7238844902153883648-LP8Y | 2024-09-09 | full 03:14.880 corporate-anniversary video, on-screen speaker cards, and platform VTT; reliability B | Six named Managing Partners narrate Advent's institutional memory. James Brocklebank, not Brooke, delivers the continuous 00:28.640–00:48.890 segment recalling Brooke's “noble experiment” wording. Exact MP4/VTT hashes preserve version identity; the promotional retrospective does not establish 1984 ownership, capital, governance, or deal credit.
• ★★ | Jacqui Morby: The Queen of Software | https://colossus.com/episode/the-queen-of-software/ | 2025-02-24 | full publisher podcast audio and show notes; reliability B | The exact MP3 is 39,000,262 bytes, runs 40:06.478, and has SHA-256 53a8da6ce20db4f99804851a8922f47b8798f78ffdef6da392851e5a8b6bccda; the RSS duration says 40:36 and the account-gated transcript was not used. Morby attributes TA's outbound software-sourcing workflow to herself, investing authority to Kevin Landry, and Brooke's role chiefly to hiring, autonomy, and culture.
• ★★★ | The venture capital industry's troubled psyche | https://fraser.stlouisfed.org/docs/publications/cfc/cfc_19750616.pdf | 1975-06-16 | contemporaneous trade-press scan; reliability B with direct quotation | John Thackray article starts printed p. 5 and Brooke quotation continues on printed/PDF p. 6, supplying rare contemporaneous direct voice on the industry's condition.

Oral histories and archives

• ★★★ | Oral History of Peter A. Brooke | https://archive.computerhistory.org/resources/access/text/2019/03/102740499-05-01-acc.pdf | interviews 2010-04-06 and 2010-09-01 | 69-page CHM/NVCA transcript, reference X8628.208; reliability A for Brooke's account | Core full-text source for chronology, vehicles, deals, Japan failure, succession, and the Cooksey dispute; printed pp. 1–63 provide stable locators.
• ★★★ | Peter Brooke Concord Oral History Program | https://concordlibrary.org/special-collections/oral-history/Brooke | 1994-10-22 | transcript and MP3 under one archive record; reliability A for Brooke's account | Age-65 interview by Renee Garrelick supplies earlier recollections, local context, TA date variants, exit-market drought, and a cross-check against the 2010 transcript.
• ★★ | NVCA Venture Capital Oral History Collection | https://computerhistory.org/press-releases/nvca-collection/ | collection announcement 2019 | archive collection metadata; reliability A for provenance | Establishes CHM custody and collection scope and routes to peer interviews; it does not independently validate interview claims.
• ★★★ | Oral History of T. Bondurant French | https://archive.computerhistory.org/resources/access/text/2021/09/102792280-05-01-acc.pdf | 2018 | CHM/NVCA witness transcript; reliability A for French's account | Printed pp. 59–61 discuss Brooke, Kevin Landry, TA's team model, spinouts, Advent, and the International Network Fund.
• ★★★ | Oral history interview with Constantine E. Anagnostopoulos | https://digital.sciencehistory.org/works/gl5p69o | 2012-08-06 | Science History Institute synchronized transcript and audio; reliability A for witness testimony | Passages at 01:20:51, 01:28–01:38, and 02:00:48 cover Monsanto's capital, Advent UK, Brooke, Cooksey, Advent International cofounder and board claims, country funds, and Genzyme attribution.
• ★★★ | Martin Y. Tang interviewed by Peter E. Hamilton | https://dome.mit.edu/handle/1721.3/204571 | interviews 2021-12–2022-01 | MIT catalog and linked 4.94 MB transcript; reliability A for Tang's account, direct bitstream resolved in reviewer QA | Hong Kong fund passage around printed pp. 69–72 describes the technology-transfer model and Tang's dissatisfaction with Advent and TA support; bind quotations to the downloaded transcript.
• ★★ | Paul Bancroft III: Early Bay Area Venture Capitalists | https://digicoll.lib.berkeley.edu/record/103596/files/bancroft_pete.pdf | 2010 | Bancroft Library witness transcript; reliability A for Bancroft's account | Printed p. 45 identifies Brooke's modest pre-Bancroft Bessemer role and Charlie Lea; printed p. 59 places both men in a wider early-industry network.
• ★★★ | Charles L. Lea Jr. oral history | https://digicoll.lib.berkeley.edu/record/103633/files/vcg-lea.pdf | oral history published through the Bancroft Library/NVCA collection | participant transcript; reliability A for Lea's account | “The Federal Express Story,” printed pp. 77 onward, and Tape 4 around 00:02:36–00:04:06 identify Lea as principal lead investor, report a first round near $23 million equity plus roughly $90 million bank borrowing, and say he spent five years making it work. This independently defeats personal Brooke-lead attribution.
• ★★★ | Wang Laboratories, Inc. records, 1948–1992 | https://id.lib.harvard.edu/ead/bak00209/catalog | 1948–1992 holdings | HBS Baker Library finding aid, call Mss: 6592 1948-1992 W246; reliability A for metadata | Subseries IIB Investment records, 1955–1986, routes to annual reports, prospectuses, shareholder letters, and Wang Family Trust interests in Tucker, TA, and Advent vehicles.
• ★★ | Peter Brooke book-development project | https://winthropgroup.com/client-portfolio/detail/peter-brooke-advent-international | project completed before 2009 publication | editorial project record; reliability B | Documents Winthrop's research and developmental-writing role, a necessary authorship boundary for the coauthored memoir.
• ★★ | Alumni Achievement Awards | https://www.alumni.hbs.edu/news-insights/alumni-achievement-awards | accessed 2026-08-02 | current HBS recipient register; reliability A for award and degree metadata | The 2000 table identifies Peter A. Brooke, MBA 1954; it confirms the honor and degree only, not the investment biography or returns.

Books and scholarship

• ★★★ | Peter Brooke chapter in Done Deals | https://archive.org/details/donedealsventure00gupt/page/244/mode/2up | 2000 | controlled-borrowing interview book; reliability A for attributed interview | Udayan Gupta chapter opens around printed p. 245; controlled access prevented full processing, so quotations and page claims remain a priority retrieval.
• ★★★ | Kevin Landry chapter in Done Deals | https://archive.org/details/donedealsventure00gupt/page/258/mode/2up | 2000 | controlled-borrowing interview book; reliability A for attributed interview | Adjacent witness chapter is essential for testing who transformed TA into its later growth-equity model and for countering Brooke-wide credit.
• ★★★ | Financiers of Innovation | https://creativeclass.com/articles/Financiers%20of%20Innovation.pdf | manuscript 1997; book 1998 | interview-based scholarly manuscript; reliability B | Complete Brooke target printed pp. 114–123 and endnotes 154–168 on pp. 354–355 rely heavily on 1987 and 1993 Brooke interviews while supplying fund-size, network, partner, and Aspen operating-example leads.
• ★★ | Venture Capital: An American History | https://books.google.com/books?id=uSmWDwAAQBAJ | 2019 | scholarly book with limited preview; reliability B | Tom Nicholas supplies long-run institutional context and comparison points; preview access is insufficient for unverified Brooke-specific claims.
• ★★ | The Advent of Venture Capital in Latin America | https://books.google.com/books/about/The_Advent_of_Venture_Capital_in_Latin_A.html?id=dPG3AAAAIAAJ | 1997 | restricted HBS case metadata; reliability C | Elizabeth B. Stein's 20-page case is a high-value route to Advent's regional model, but full text and exhibits were not accessible.
• ★★ | Venture Capital and High Technology Entrepreneurship | https://kenney.faculty.ucdavis.edu/wp-content/uploads/sites/332/2018/03/Venture-Capital-and-High-Technology-Entrepreneurship.pdf | 1988 | scholarly article manuscript; reliability B | Regional-history comparator contains Brooke and Bank of Boston claims but compresses dates, requiring control against direct and contemporaneous records.
• ★★★ | Une fabrique publique d’entrepreneurs schumpétériens en France (1971–1978) | https://journals.openedition.org/nrt/13460 | 2023 | peer-reviewed archival scholarship; reliability B | Jean-Baptiste Devaux assigns the 1971 innovation white paper to Christian Marbach and Thierry Gaudin, describes Arthur D. Little as making Boston introductions, and records Marbach's meeting with Brooke; this complicates, but does not disprove, Brooke's claim of a separate subcontracted venture-capital section.
• ★★ | Translating emergent technologies into novel therapeutics | https://orca.cardiff.ac.uk/id/eprint/179347/1/translating-emergent-technologies-into-novel-therapeutics-tracing-complementarity-and-co-evolution-in-the-cambridge-boston-innovation-ecosystem.pdf | 2025 | scholarly study; reliability B | Printed p. 203 note 71 / PDF p. 18 cites a 1980 Biogen shareholder letter identifying TA Associates as an investor. It is a route/control for TA participation, not the original letter, Brooke decision evidence, or a return record.

Press

• ★★★ | Peter Brooke, dean of Boston's venture capitalists, dies at 90 | https://www.bostonglobe.com/2020/04/07/metro/peter-brooke-dean-bostons-venture-capitalists-dies-90/ | 2020-04-07 | reported obituary; reliability B | Bryan Marquard gives Peter Albert Brooke's identity, witnesses, a 1995 CEO handoff, Wang governance, and a reported Biogen gain that remains unaudited.
• ★★ | Peter Brooke obituary | https://www.legacy.com/us/obituaries/bostonglobe/name/peter-brooke-obituary?id=2246273 | 2020-04 | family obituary; reliability B | Family, education, Army, TA, Advent, chair, and death chronology; interested and derivative for investment claims.
• ★★ | TA Associates reflects on the life of founder Peter Brooke | https://www.ta.com/news/ta-associates-reflects-on-the-life-of-founder-peter-brooke/ | 2020-04-10 | firm memorial; reliability B | Official TA founder and succession narrative, useful for dates and witnesses but promotional and not an independent performance source.
• ★★ | Peter Brooke memorial biography | https://www.bpea-pe.com/team/peter-brooke/ | accessed 2026-08-02 | successor-firm biography; reliability B | Concise 1961, 1963, 1966, 1968, 1973, 1984, 1996, and 2015 chronology plus education; interested institutional family.
• ★★ | Brook announces that John Brooke is appointed as an advisor to Brook II | https://www.brookventure.com/brook-announces-that-john-brooke-is-appointed-as-an-advisor-to-brook-ii/ | 2006-02-09 | corporate announcement; reliability B | Paragraph beginning “John and Peter Brooke” reports BPEA's 2002 founding and the Brooke Family Limited Partnership's 1995 predecessor date; interested and not formation-record proof.
• ★★ | Peter and Anne Brooke give collection to Harvard Art Museums | https://news.harvard.edu/gazette/story/2007/09/peter-and-anne-brooke-give-collection-to-huam/ | 2007-09 | university news; reliability B | Contemporary institutional affiliation, philanthropy, and family lead; does not validate venture returns.
• ★★★ | Venture capital firm leads way to new kinds of lending | https://www.washingtonpost.com/archive/business/1990/03/07/venture-capital-firm-leads-way-to-new-kinds-of-lending/013f1628-b944-4bb3-b6d6-edb3260ca10d/ | 1990-03-07 | contemporaneous business report; reliability B | David Warsh documents TA's shift and Kevin Landry's role, providing an external check on founder-centered institutional history.
• ★★ | Adams Russell electronics spin-off report | https://www.worldradiohistory.com/Archive-All-BC/Broadcasting-Magazine/BC-1986/BC-1986-06-30.pdf | 1986-06-30 | contemporaneous trade-press scan; reliability B | Reports the board-approved separation/spin-off sequence for Adams Russell's electronics operations; company chronology does not prove Brooke's causal role, ownership, or proceeds.
• ★★ | Adams Russell cable transaction report | https://www.worldradiohistory.com/Archive-All-BC/Broadcasting-Magazine/BC-1987/BC-1987-07-27.pdf | 1987-07-27 | contemporaneous trade-press scan; reliability B | Reports the later cable-system transaction path. Use with Brooke's retrospective proposal account; do not convert the company transaction into a Brooke return.
• ★★ | Venture's New Grail | https://www.forbes.com/forbes/2009/0216/046.html | 2009-01-29 | business feature; reliability B | Later account of TA outbound sourcing and growth strategy centers team and Landry; useful but retrospective.
• ★ | Selfless endeavor | https://www.forbes.com/global/2000/0221/0304064a.html | 2000-02-21 | business feature; reliability B | Context for Brooke's Endeavor advisory role; peripheral to portfolio performance.
• ★★★ | Un fondo internacional de capital-riesgo invertirá 25.000 millones de pesetas en empresas españolas | https://elpais.com/diario/1988/02/28/economia/573001203_850215.html | 1988-02-28 | contemporaneous Spanish report; reliability B | Identifies the International Network Fund's reported scale and Spanish participants, useful for entity and geography reconstruction.
• ★★ | Advent International raises $10.8 billion buyout fund | https://www.bostonglobe.com/business/2012/11/15/advent-international-boston-private-equity-firm-raises-bilion-buyout-fund/qmWJo2iwoiV8oC3jNawzIK/story.html | 2012-11-15 | contemporaneous business report; reliability B | Later-firm scale and Brooke remarks, to be separated from his operating-era funds and personal track record.
• ★★ | Kevin Landry, Boston private equity executive, dies | https://www.bostonglobe.com/business/2013/08/05/kevin-landry-boston-private-equity-executive-known-for-his-personality-and-zeal-for-work-and-life/LlQzEMDjV0BUVFFjQlevZL/story.html | 2013-08-05 | reported obituary; reliability B | Independent witness and chronology route for Landry's TA leadership and the Brooke-to-Landry attribution boundary.
• ★ | Clouds over the VC sky | https://bostonbiotechwatch.com/tag/private-equity/ | 2009-10-26 | attendee recap; reliability C | Partial recap of the Brooke–McGuire forum and fundraising pessimism; locate by title on the tag page and do not quote as a transcript.

Deal data and filings

• ★★★ | Devonshire Capital Corp. application for an SBIC license | https://www.govinfo.gov/content/pkg/FR-1975-09-16/pdf/FR-1975-09-16.pdf | 1975-09-16 | Federal Register notice; reliability A | Printed p. 42805 gives $1.25 million capitalization, Advent II and Bermuda ownership, Boston University stake, TA advisory role, and named officers excluding Brooke.
• ★★★ | Advent IV Capital Co. application for an SBIC license | https://www.govinfo.gov/content/pkg/FR-1981-10-26/pdf/FR-1981-10-26.pdf | 1981-10-26 | Federal Register notice; reliability A | Printed p. 52267 gives $6.025 million capitalization, SBIPCo. as general partner, and Brooke, Landry, Croll, and Stamps as partners of that general partner.
• ★★★ | Excelsior Buyout Investors prospectus | https://www.sec.gov/Archives/edgar/data/1213980/000119312504100166/d486bpos.htm | 2004-06-08 | SEC-filed prospectus; reliability A for filing content | Underlying-fund Advent section gives a 1984 TA spinout narrative, reported capital raised since 1985, strategy, and later organization; claims originate with fund materials.
• ★★★ | Aspen Technology final prospectus | https://www.sec.gov/Archives/edgar/data/929940/000095013596002588/0000950135-96-002588.txt | 1996-06-11 | SEC-filed 424B4; reliability A | Management section says Douglas Brown had been a director since 1986 and became president and CEO in January 1996, a succession control against Brooke's 1997 recollection.
• ★★★ | Unitrode Corporation 1995 proxy statement | https://www.sec.gov/Archives/edgar/data/101911/000095013595001016/0000950135-95-001016.txt | filed 1995-04-28 | SEC-filed DEF 14A; reliability A | Election-of-directors table identifies Peter A. Brooke as a director since 1962, Advent chair, and NEBS director; securities-ownership table as of 1995-04-14 reports 36,000 beneficial Unitrode shares, less than 1%, including options exercisable for 20,000. This is observed point-in-time ownership, not an exit holding or return.
• ★★★ | Texas Instruments–Unitrode merger registration statement | https://investor.ti.com/sec-filings/sec-filing/s-4/0000950134-99-008312 | filed 1999-09 | TI-hosted SEC Form S-4, accession 0000950134-99-008312; reliability A for filing content | Merger materials set the stock consideration and use a maximum 35,511,855 Unitrode shares with a 1999-09-09 reference price of $44.28; TI's later company record says the acquisition closed in October 1999. The filing does not prove Brooke retained his 1995 position through closing.
• ★★★ | Federal Express financing chronology | https://law.justia.com/cases/federal/district-courts/FSupp/645/1281/1499387/ | 1986 federal district-court opinion recounting 1973–74 financings | adjudicative record transcription; reliability A for the court record | Describes the May 1973 aircraft financing, November 1973 $24.5 million financing from 24 investors, and March/September 1974 follow-ons. It does not identify Brooke as lead or supply TA's check, stake, or proceeds.
• ★★★ | Deluxe acquisition of New England Business Service | https://www.sec.gov/Archives/edgar/data/27996/000089710105000730/deluxe050979_10k.htm | filed 2005 | SEC-filed Deluxe Form 10-K; reliability A | Note 4 says Deluxe acquired all outstanding NEBS common stock on 2004-06-25 for $44 per share and records total purchase price of $639.789 million. It establishes company exit terms, not Brooke's holding or proceeds.
• ★★★ | Corning acquisition of Damon | https://www.sec.gov/Archives/edgar/data/24741/0000950146-94-000007.txt | filed 1994-01-18 | SEC-filed Corning Form S-3/A, accession 0000950146-94-000007; reliability A for filing content | “Acquisition of Damon” says Corning acquired all outstanding Damon shares on 1993-08-04 for approximately $405 million including expenses and refinanced about $167 million of Damon debt. It does not establish continuity from Brooke's early financing to the exit.
• ★★★ | Damon Corporation proposed FTC consent order | https://www.govinfo.gov/content/pkg/FR-1977-07-11/pdf/FR-1977-07-11.pdf | 1977-07-11 | Federal Register notice, 42 FR 35660; reliability A for agency record and allegations | Printed p. 35660 describes proposed acquisition restrictions and alleges physician-referral payment practices from at least 1971 to mid-1975. Treat allegations/order history as company-level counterevidence, not proof of Brooke's participation.
• ★★★ | Damon Corporation FTC order-modification decision | https://www.ftc.gov/sites/default/files/documents/commission_decision_volumes/volume-101/ftc_volume_decision_101_january_-_june_1983pages_687-839.pdf | 1983-03-29 | official FTC decision volume, 101 F.T.C. 689; reliability A | Printed pp. 689–694 contain the order-to-show-cause disposition and competing commissioner views on acquisition restrictions and competitive effects. Use to qualify Damon's growth-by-acquisition narrative; it is not personal Brooke misconduct evidence.
• ★★★ | Terapia sustainable private-equity case study | https://documents1.worldbank.org/curated/en/456971468324046031/pdf/382170Promise01ity0Nov200601PUBLIC1.pdf | 2006 | IFC/World Bank case-study publication; reliability A for IFC's institutional account | Printed pp. 32 onward identify Advent's 2003 $49.5 million leveraged public-to-private transaction and describe environmental, governance, and operating work. The transaction is successor-era and must not enter Brooke's operating track record.
• ★★ | Aspen Technology founder presentation | https://cache.org/sites/default/files/Evans%20-%20CACHE%20Presentation%2008-08-09.pdf | 2009-08 | founder/company-history presentation; reliability B for retrospective participant account | AspenTech history slides report that the company, dangerously short of cash, raised $2.6 million from Advent in 1986. Use with the 1996 SEC prospectus; the slides do not identify vehicle ownership or realized proceeds.
• ★★ | Kevin Landry board-history control | https://www.sec.gov/Archives/edgar/data/1283699/000095013407008244/d42548a3exv99w1.htm | filed 2007 | SEC-filed MetroPCS exhibit; reliability A for filed biography | Identifies Kevin Landry as a former director of Biogen and Continental Cablevision. This supports team-attribution caution and does not by itself prove the investing vehicle, check, or return.
• ★★ | Vectis Healthcare & Life Sciences Fund II Form D | https://www.sec.gov/Archives/edgar/data/1444009/000144400909000002/xslFormDX01/primary_doc.xml | filed 2009 | SEC Form D; reliability A for filed notice, subject to the SEC's non-review warning | Identifies Vectis II as a Delaware limited partnership organized in 2008, lists the Brooke Private Equity Advisors address, and names Vectis II GP, L.P. as promoter. This is a later BPEA entity-control row, not Brooke-era TA/Advent performance.
• ★★ | ILPA Principles 3.0 | https://ilpa.org/wp-content/uploads/2019/06/ILPA-Principles-3.0_2019.pdf | current published principles accessed 2026-08-18 | Institutional Limited Partners Association best-practice guidance; reliability A for ILPA's recommendations | Printed pp. 10–22 and glossary p. 41 provide a current LP-fund comparison for economics, term extensions, conflicts, key-person provisions, GP ownership and definitions. These are recommendations, not law or a measured market average, and are not imputed to Brooke-era agreements.
• ★★★ | Advent International Limited company overview | https://find-and-update.company-information.service.gov.uk/company/02342186 | incorporated 1989-02-03 | UK Companies House record; reliability A | Establishes the UK entity's company number, incorporation, status, and filings; it is not proof that the global firm began in 1989.
• ★★★ | Advent International Limited officers | https://find-and-update.company-information.service.gov.uk/company/02342186/officers | officer records from 1989 onward | UK Companies House register; reliability A | Identifies Peter Albert Brooke and role dates, providing identity and UK-precursor controls distinct from U.S. Advent's 1984 or 1985 formation.
• ★★ | About Advent International | https://www.adventinternational.com/about-us/ | accessed 2026-08-02 | current firm history; reliability B | Official 1984 founding, team, global scale, and strategy; promotional, drift-prone, and not an early fund ledger.
• ★★ | Advent investments | https://www.adventinternational.com/investments/ | accessed 2026-08-02 | current selected portfolio; reliability B | Candidate company enumeration only; may omit losses, exited legacy investments, local affiliate deals, and investments outside the displayed strategy.
• ★★ | About TA | https://www.ta.com/about/ | accessed 2026-08-02 | current firm history; reliability B | Firm-reported strategy and scale, useful for lineage but not Brooke-era economics or individual attribution.
• ★★ | TA portfolio | https://www.ta.com/portfolio/ | accessed 2026-08-02 | current selected portfolio; reliability B | Candidate company list requiring vintage, vehicle, partner, loss, and exit reconciliation before inclusion in a Brooke denominator.
• ★★ | BPEA portfolio | https://www.bpea-pe.com/portfolio/ | accessed 2026-08-02 | current successor-firm portfolio; reliability B | Separates later BPEA activity from TA and Advent; selected-list and attribution limits remain.
• ★★ | SEAVI Advent | https://www.seaviadvent.com/ | accessed 2026-08-02 | current regional firm site; reliability B | Route to Southeast Asian entity lineage, team, and candidate portfolio; do not assume common ownership or Brooke selection across all vintages.
• ★★ | New Hampshire Retirement System investment committee materials | https://www.nhrs.org/docs/default-source/iic-public-materials/iic-public-materials---june-2025.pdf?sfvrsn=b59314b4_3 | 2025-06 | public-LP meeting materials; reliability A for LP record, B for manager claims | Later Advent organization and fund-underwriting route; snapshot is far outside Brooke's operating era and not personal performance evidence.
• ★★ | Advent International GPE IX investment memorandum | https://data.treasury.ri.gov/dataset/7c2eed50-9cae-46f0-b320-34bd911d3450/resource/4be2e34a-b879-4b4e-8305-f6420f3e2f0b/download/Cliffwater-RISIC-Advent-International-GPE-IX-Memo-031819.pdf | 2019-03-18 | Rhode Island public-LP consultant memo; reliability A for dated record | Later-fund strategy and performance route; its “formed in 1994” statement conflicts with stronger founding evidence and demonstrates secondary memo error.
• ★★ | Advent International GPE X Cliffwater memorandum | https://data.treasury.ri.gov/dataset/44f0a63a-fb87-4461-a655-2f43c58eea40/resource/bbf046a8-15fd-4c46-b0a5-8e6ce375f800/download/3B-Advent-GPE-X-Cliffwater-RISIC-Memo-04152022.pdf | 2022-04-15 | Rhode Island public-LP consultant memo; reliability A for dated record | Says founded in 1984 and records later strategy, terms, and performance claims; not an audited Brooke-era denominator.
• ★★ | Advent International GPE X Rhode Island presentation | https://data.treasury.ri.gov/dataset/44f0a63a-fb87-4461-a655-2f43c58eea40/resource/d17ca5b0-f427-42fc-b68a-537fdde36fcd/download/3c-advent-international-gpe-xrhode-islandapril-272022final.pdf | 2022-04-27 | manager presentation to public LP; reliability B | Pages 3, 15, and 17 report ESSF, GPE II, and GPE III history or economics; firm-supplied and unaudited, but useful as denominator leads.
• ★★ | University of Houston Endowment Management Committee minutes | https://uhsystem.edu/board-of-regents/documents/minutes/11-19-15-Endow%20Minutes.pdf | 2015-11-19 | public-institution minutes; reliability A for decision record | Advent GPE VIII recommendation and later firm/fund assertions provide a dated LP route, not early vehicle cash flows.
• ★★ | IFC Annual Report 2001, volume 2 | https://www.ifc.org/content/dam/ifc/doc/mgrt/ar2001-english-vol2.pdf | 2001 | official institutional portfolio report; reliability A | SEAVI and regional project entries help reconstruct country vehicles and institutional capital without assigning the whole portfolio to Brooke.
• ★ | Advent ESG Overview 2023 | https://www.adventinternational.com/wp-content/uploads/2024/09/Advent_ESG-Overview_2023.pdf | 2024 | corporate retrospective; reliability B | Brooke quotation and current firm narrative are useful only as later institutional memory, not as independent chronology or performance proof.

Historical context

• ★★★ | Small Business Investment Act of 1958 | https://www.govinfo.gov/content/pkg/STATUTE-72/pdf/STATUTE-72-Pg689.pdf | 1958-08-21 | enacted federal statute; reliability A | Primary legal context for SBIC formation and financing; do not infer that it alone caused TA's vehicles or outcomes.
• ★★★ | Investment Duties final rule, 44 Federal Register 37221 | https://archives.federalregister.gov/issue_slice/1979/6/26/37221-37229.pdf | 1979-06-26 | final federal regulation; reliability A | Primary prudent-investor rule text; it changed fiduciary analysis but should not be summarized as simply “legalizing venture capital.”
• ★★ | The Marshall Plan speech | https://www.marshallfoundation.org/the-marshall-plan/speech/ | 1947-06-05 | institutional transcript of public address; reliability A | Controls the content of the Harvard speech Brooke later cited as formative; no direct causal link to his career is documented.
• ★★★ | IFC: The First Six Decades | https://documents1.worldbank.org/curated/en/668851478627391927/pdf/109806-WP-IFC-History-Book-Web-Version-OUO-9.pdf | 2016 | official institutional history; reliability B | Page 47 discusses SEFINNOVA and supplies institutional context for early cross-border venture vehicles and IFC participation.
• ★★ | The Next New Thing: A History of the Venture Capital Industry | https://computerhistory.org/stories/the-next-new-thing/ | accessed 2026-08-02 | CHM historical synthesis; reliability B | Broad early-industry and oral-history context for comparison, not an independent Brooke deal ledger.
• ★★★ | Le Cahier de l'Innovation | https://www.persee.fr/authority/610047 | 1971 | French institutional publication catalog; reliability A for metadata | SUDOC 002091496 attributes the 189-page publication to the Ministère du développement industriel et scientifique with a François-Xavier Ortoli preface; the catalog supplies no Brooke credit or body, so it is an authorship-control route rather than Brooke prose.

Criticism and skeptical evidence

• ★★★ | Venture capital companies: an evaluation of eight early funds | https://documents.worldbank.org/en/publication/documents-reports/documentdetail/870201468156255238 | 1993 | official IFC evaluation; reliability A | Evaluates eight venture-capital companies launched from 1976–1986 and reports mixed or poor early results, a direct control against frictionless global diffusion claims.
• ★★★ | Venture capital funds in emerging markets: lessons from IFC's investments | https://documents.worldbank.org/en/publication/documents-reports/documentdetail/166371468782154146 | 1996 | official IFC Note 90; reliability A | Barger, Carter, and Kuczynski identify structural and performance problems in emerging-market funds; context, not proof about every Advent affiliate.
• ★★ | Two and Twenty: Taxing Partnership Profits in Private Equity Funds | https://papers.ssrn.com/sol3/papers.cfm?abstract_id=892440 | 2006 | legal scholarship preprint; reliability B | Victor Fleischer's carried-interest critique supplies the framework behind his WBUR counterpoint and separates tax advantage from investing skill.
• ★★★ | Wang Laboratories files for bankruptcy | https://www.nytimes.com/1992/08/19/business/wang-laboratories-files-for-bankruptcy.html | 1992-08-19 | contemporaneous business report; reliability B | Chapter 11 control for Brooke's long board relationship and succession narrative; bankruptcy is not by itself proof of his personal causation.
• ★★ | Private Equity: Recent Growth in Leveraged Buyouts Exposed Risks That Warrant Continued Attention | https://www.gao.gov/products/gao-08-885 | 2008-09-09 | U.S. government report; reliability A | Sector-level leverage, transparency, and oversight critique contemporaneous with Brooke's later advocacy; mostly post-operating-era context.
• ★ | Venture Capital's Early Pioneers | https://www.institutionalinvestor.com/article/2btgdtwr5w5i99g1zmm0w/portfolio/venture-capitals-early-pioneers | 2010-02-06 | trade-press history; reliability B | Interview-derived pioneer narrative contains compressed entity and chronology claims to test rather than inherit.
• ★★ | Sandvik AB v. Advent International Corp. | https://law.justia.com/cases/federal/appellate-courts/F3/220/99/625483/ | 2000 | federal appellate opinion transcription; reliability A for the adjudicative record | Records a post-Brooke dispute over affiliate authority and contract formation; the court did not decide the underlying fraud or contract merits, and the case is not personal evidence about Brooke.

Coverage assessment

• Source-map protocol: every row remains [ ] because SCOUT maps and inspects access without claiming later source-level processing. ★★★ is the highest retrieval queue, followed by ★★ and ★. Reliability concerns both provenance and the exact claim: a filing is authoritative for what was filed, not necessarily for manager-supplied history embedded inside it.
• Corpus shape: this map contains 88 source rows and 88 unique retrieval URLs across direct speech, Brooke oral histories and claimed writings, witness testimony, archives, books, contemporaneous press, federal and corporate records, public-LP documents, scholarship, and skeptical institutional evaluations. Institutional clusters—especially CHM/NVCA/Bancroft, HBS, TA, Advent/BPEA, WBUR, Rhode Island, IFC, the Unitrode/TI and Damon/FTC transaction families, the French ministry/Devaux record, and the Boston Globe—are not independent merely because they have separate URLs; roughly 68 source families remain after those clusters and the two Done Deals chapters are reconciled.
• Priority and category totals: 41 ★★★, 43 ★★, and 4 ★ items. Categories contain 4 writings; 6 talks/video or contemporaneous direct-voice items; 11 oral-history/archive items; 8 books/scholarship items; 15 press items; 31 deal/filing and fund-governance items; 6 historical-context items; and 7 criticism/skeptical items.
• Identity and namesake sweep: Peter A. Brooke, Peter Albert Brooke, Peter Brooke, P. A. Brooke, and the unsupported Peter Adams Brooke variant were paired with Bank of Boston, Bessemer, Tucker Anthony, TA Associates, Advent, Sofinnova, SEAVI, Wang, Unitrode, Genzyme, Biogen, and partner names. Sir Peter Brooke the British politician, the South African Old Mutual executive, the painter, Peter Brooke Scott, and unrelated engineering or academic namesakes were excluded.
• Attempted-route ledger: direct voice was searched through CHM/NVCA, Concord, WBUR and its Megaphone assets, HBS, C-SPAN, YouTube, Colossus/RSS, podcasts, LinkedIn's public VideoObject/MP4/VTT routes, Xconomy and WilmerHale event archives, Common Crawl historical indexes and byte-range ARC/WARC retrieval, Boston Biotech Watch, and exact-title searches; writings through exact titles and variants, JSTOR, HeinOnline, ABA, Government of Canada, FGV, CODIMEC/CODEMEC, ACRJ, FIESP, publishers, library catalogs, controlled lending, Google Books, Internet Archive, HathiTrust-style routes, repository searches, and Arthur D. Little or bank-paper variants; French authorship through SUDOC, Persée, OpenEdition, archive call numbers, ministry names, Marbach, and Gaudin; vehicles and deals through GovInfo, SEC, Companies House, HBS, IFC/World Bank, public-LP records, firm sites, country names, and company or partner combinations; criticism through bankruptcy, failure, litigation, sanction, carried-interest, regulation, leverage, and dissent terms. The bounded sweep found no public congressional testimony, LP-letter corpus, standalone papers archive, audited early fund series, substantiated personal sanction, or authenticated Brooke–McGuire forum audio/transcript.
• Direct voice and access: the 69-page CHM/NVCA transcript and 1994 Concord transcript plus audio are the richest Brooke oral histories. WBUR's 2007 exact MP3 runs 43:43.216; its 2009 live MP3 runs 46:41.946 while the page displays 24:22, so every WBUR locator binds to a hashed file version. Morby's full 40:06.478 publisher MP3 is accessible, while the account-gated transcript was not used. Advent's full 03:14.880 LinkedIn video and VTT preserve named later-witness testimony but no Brooke recording. For the Brooke–McGuire family, three complete Xconomy article bodies and a named attendee recap were recovered; the separate 2009-09-23 Brooke interview is not October 21 stage testimony, the priority-appearance denominator remains 11, and authenticated forum audio/transcript is unavailable. Of the recovered written record, only the 1975 trade-press article and the bounded Florida/Kenney manuscript section are fully readable; the 2009 book is partial-only; the 1971 and 1973 articles, 1981 CODIMEC contribution, reported bank paper and French study, and both Done Deals chapters lack a lawful body. No public corpus of LP letters, investment memoranda, correspondence, or standalone Brooke papers surfaced.
• Archive routes: CHM reference X8628.208, Concord Special Collections, HBS Wang call Mss: 6592 1948-1992 W246 subseries IIB, Bank of America Corporate Archives' restricted FleetBoston/acquired-institution holdings, MIT Arthur D. Little records MC-0579, Science History Institute audio/transcript, MIT's Tang record, Berkeley's Bancroft transcript, French economic archive SAEF 1A127/3, and Archives nationales 19820043/1 are actionable. MIT exposes the Tang catalog and linked 4.94 MB transcript; the direct bitstream resolved, and printed pp. 69–72 are fully analyzed with claim-local citations in the TALKS corpus. The Bank policy paper, a distinct Arthur D. Little Route 128 report, original CODIMEC proceedings, Bank credit files, Tucker/TA formation papers, Bessemer files, early LP statements, and the cited French archive files remain missing or inaccessible.
• Chronology conflicts: preserve Bank portfolio $10 million versus $20 million variants; Wang's reported $150,000 facility versus $50,000 plus a $100,000 credit line; TA activity or formation in 1966, 1967, 1968, or 1971; Sofinnova in 1972 or 1973; Advent in 1984 or 1985; the UK entity as a later or precursor legal route rather than the global founding date; and the CEO handoff in 1995, 1996, or 1997. Aspen's filing places Douglas Brown in the CEO role in January 1996. Genzyme vehicle/date claims and any “first global” claim remain provisional.
• Vehicle and attribution boundary: The corpus exposes 25 named vehicle/mandate rows or families; 25/25 were searched, 4 support only narrow hard capitalization or reported fund-economics claims, 21 remain retrieval-only for economics, and 0 supports a Brooke-personal return. Bank lending, Bessemer family capital, Tucker activity, each TA partnership/SBIC, Advent management companies, country funds, corporate accounts, INF, ESSF/GPE funds, SEAVI, and BPEA/Vectis remain separate. Lea identifies himself as principal FedEx lead; a filed biography places Landry on Biogen and Continental boards; Morby reserves her sourcing method to herself/Landry/team. Researcher inference · strong for the separation and denominator; source-level testimony/filings retain their own grades.
• Completed-peer comparison: Georges Doriot operated through publicly traded ARD; Laurance Rockefeller deployed inherited family capital through a family investment platform; and Reid Dennis moved from personal and insurer capital into limited partnerships. Brooke instead moved from bank lending through a family investment office into successive partnerships and an international affiliate network. Later phases should compare sourcing channels, ownership, governance, incentives, portfolio construction, and failures using each peer's underlying evidence; this is a vehicle-and-access scaffold, not a ranking or proof that any person was first.
• Candidate denominator: 17/17 publicly named company/opportunity candidates were searched and entered; 0/17 has a complete Brooke-attributable entry, ownership, follow-on, decision-rights, and exit-cash-flow reconstruction. Eleven are core Brooke-linked qualitative rows and six are era/team controls. Bank/Japan/INF/ChinaVest/Hong Kong/anonymous-failure evidence remains separately counted program context, not silently added as companies. No audited early TA/Advent portfolio, complete loss/pass log, fee/carry schedule, or Brooke-personal denominator surfaced; firm-supplied GPE performance remains unaudited. Researcher inference · strong.
• Critical and negative coverage: Brooke directly acknowledges a long 1970s realization drought, Wang's collapse, failure of the affiliate model in Japan, and limited technology success outside the U.S. except Israel. Tang describes insufficient support for a Hong Kong concept; IFC reports mixed early emerging-market VC outcomes; Cooksey disputes Brooke's account of his departure; and Landry/team sources resist founder-wide attribution. A Sandvik court matter surfaced only as a post-Brooke institutional-control lead and is not evidence about Brooke personally. No substantiated personal criminal conviction or venture-related regulatory sanction surfaced in this bounded sweep, but that result is not proof of absence.
• Highest-value retrievals: (1) Bank policy/credit-committee and borrower files; (2) the 1971 Business Lawyer and 1973 Pensions bodies; (3) original CODIMEC proceedings and the alleged ADL/French archive bodies; (4) TA/Advent partnership, LP, IC, allocation and cash-flow records; (5) Wang HBS materials; (6) Unitrode's 1961 subscription/basis and 1996–99 Forms 4/5 or holder continuity; (7) Biogen's original TA memo/capitalization/distributions beyond the BHR/Landry controls; (8) Aspen's 1986 subscription/cap table beyond the founder presentation and 424B4; (9) Genzyme, Federal Express, Damon, NEBS and Adams Russell entry-to-exit holding records; and (10) complete pass, write-off, reserve and litigation ledgers beyond Morby's partial set.

Source notes