Active Portfolio Management: Rates, Liquidity, and Tools
Panel discussion · speakers attributed by role
Headline takeaway
A practitioner panel on running the liquid book argued that with cash yielding well, the justification for active risk is income and downside management, not beating the index by a point or two. The throughline: harvest volatility premium through options and structured notes, pre-commit to a rules-based rebalancing policy so you buy into drawdowns rather than freeze, and treat the now-ubiquitous endowment model as a live illiquidity risk rather than a settled answer. The recurring family-office edge they kept returning to is time — the freedom not to play.
Key points
- On why take active liquid risk at all when cash pays: the panel's answer was income enhancement. Their method is selling volatility through options and structured notes, harvesting the implied-vol premium as (often more tax-advantaged) income, and layering those structures on top of long-term holdings so the two return streams smooth each other out. Crypto-linked equities were singled out as a place to harvest a large premium, with implied vol on bitcoin and some US names running north of 80% in a FOMO market — provided you are willing to own the underlying.
- On build-versus-buy, the panel did both but leaned toward humility: as a small shop, allocate the specialist strategies to external managers you trust to execute better than you can. Structured-note pricing doubles as intelligence — when one bank will "trade wide" (a panelist cited a European desk pricing about 25bps wider than peers), it tells you something about that counterparty and the health of the financial sector.
- A pointed view on internal product teams: if a family office builds a genuinely hot trading or product team, question whether it belongs there long term, because it biases the whole operation. A true family office is custody of capital across the entire portfolio — managing psychology and generations — not a product shop; if the team is that good, spin it out as a separate entity (be the anchor LP and take outside capital alongside).
- On positioning for the next ten years, the mood was cautious. Rates are higher, central-bank balance sheets are no longer expanding, geopolitical risk is more visible, and several speakers the panel had heard recently are flagging much more muted forward equity returns on current valuations. The hardest variable is AI capex: the panel's view is that you cannot judge the return on the spend until roughly 18 months after it lands, and floated "Anthropic at a trillion dollars of revenue in 4-5 years" as the kind of upside no one can yet underwrite.
- A favourite framing was the concentration of equity returns: across US companies from 1900 to 2016, about 4% generated all of the net return, roughly 55-57% merely matched risk-free, and the rest lost money. The job is finding the 4%.
- A specific alpha idea: mag-7 capex is reaching around $900bn and flows heavily into materials and commodities, yet index exposure to materials and commodities is very low — so overweighting them may be a source of alpha. High inflation, they noted, tends to favour equities.
- On liquidity, a former CIO who acted as a liquidity provider through COVID made the structural case: you cannot time drawdowns, so the discipline has to be pre-set. Their rebalancing policy forces a shift out of treasuries into equities at a 20% drawdown, precisely because in the middle of a sell-off no one wants to hit the buy button. Set the rule when level-headed, not in the crisis.
- The endowment model drew the most scrutiny. Everyone has been in roughly the same illiquid trade for 25 years since Swensen, and the model cannot rebalance: you commit when markets are high and end up over-weighted in illiquid assets when they fall. The panel's worked examples were the GFC — Harvard selling at discounts and borrowing about $1bn to meet capital calls, Yale levering through, and a teachers' pension forced into a swap with JPMorgan to pay pensions because it was so illiquid. With the model now pervasive even among family offices, the panel argued the jury is still out versus simply owning public markets.
Notable claims, calls, or numbers
- Implied vol on bitcoin and some US names running north of 80% in FOMO markets — the raw material for income harvesting, if you will hold the underlying.
- US equity-return concentration, 1900-2016: ~4% of companies generated all the net return; ~55-57% matched risk-free; the rest were net losers.
- Mag-7 capex approaching $900bn, flowing into materials and commodities that are under-represented in indices — flagged as a potential alpha source.
- Rules-based rebalancing trigger: rotate treasuries into equities at a 20% drawdown.
- Endowment-model stress cases: Harvard borrowed ~$1bn in the GFC to fund capital calls; a teachers' pension did a JPMorgan swap to pay pensions; S&P annualised ~15% over the period while a panelist claimed KKR has not had a fund return 2x since 2012.
- Active-management hurdle: outperforming the S&P 500 at all is admirable; ~100bps of alpha is a reasonable bar; one panelist with a 20-year record called 7% a great number and would take an extra 1% a year for the risk management alone.
- FX as a driver: Microsoft earns about 50% of revenue in foreign currency; Rolls-Royce ran roughly 10x in 12 months as an example of single-name dispersion.
- Cost of leverage on the untouched core book cited at 4-5%; tail-risk funds that pay roughly 3x when vol spikes.
Disagreements or tensions
- The most honest tension was active versus passive, surfaced rather than averaged: a panelist recounted doing all the work to beat the index by 1-2% and his principal asking whether they could have just owned the index and gone to the beach. The counter is that alpha is not only upside — it is downside capture and risk management, and part of what you are buying is peace of mind from someone actively protecting the capital.
- On the endowment model there was real skepticism rather than consensus: it has proliferated into family offices, but the panel was unconvinced it beats public-market constructs once you price in the illiquidity that bit endowments in the GFC.
- On internal capability there was a clear split between the information edge of being in the market yourself and the bias a strong in-house product team introduces into a family office that is supposed to steward the whole portfolio.
Implications for portfolio positioning
- Build the liquid book around income and protection rather than index-beating: vol-premium harvesting via options and structured notes, sized to a willingness to own the underlying, with tail hedges that pay multiples in a spike.
- Pre-commit liquidity and rebalancing rules in calm conditions — a hard drawdown trigger to deploy treasuries into equities — so behaviour is automatic when markets are falling.
- Treat illiquid and endowment-style allocations as a forced-seller risk: size private exposure so capital calls and drawdowns never make you the one selling at a discount.
- Hold AI-capex humility on ten-year equity assumptions; the materials-and-commodities-versus-index-weight gap is a concrete, named place to look for alpha given roughly $900bn of mag-7 capex.
- For non-USD families, FX is a first-order return driver — run an explicit hedging policy rather than a default.
- Operationally, choose a custodian you trust to be there in a crisis, monitor counterparty CDS (especially for structured products), use structured-note pricing dispersion as a counterparty-health signal, and be wary of what private banks and brokers push.
- With cash yielding well, the justification for active risk is income and downside management, not beating the index by a point or two: harvest volatility premium through options and structured notes (implied vol north of 80% on bitcoin and some US names), as long as you are willing to own the underlying.
- Pre-commit a rules-based rebalancing policy — for example, rotate treasuries into equities at a 20% drawdown — because in the middle of a sell-off no one wants to hit the buy button. The family-office edge is time, the freedom not to play, paired with tail hedges that pay ~3x on a vol spike.
- Treat the endowment model as a live illiquidity risk, not a settled answer: everyone has been in the same trade for 25 years, and the GFC caught it (Harvard borrowed ~$1bn for capital calls, a teachers' pension swapped with JPMorgan), while the S&P compounded ~15% and KKR has not returned 2x on a fund since 2012.
- Hold AI-capex humility on ten-year returns — you cannot judge the spend for ~18 months — and note a named alpha source: heavy mag-7 capex (~$900bn) flows into materials and commodities that carry very low index weight. For non-USD families, FX is first-order (Microsoft earns ~50% of revenue abroad).
- Operationally: choose a custodian you trust in a crisis, monitor counterparty CDS, use structured-note pricing dispersion as a counterparty-health signal, and be wary of what private banks push. About 100bps of alpha is a reasonable bar; the honest tension is whether the whole apparatus beats simply owning the index.
How Families Deploy AI Within Their Workflow
Panel discussion · speakers attributed by role
Headline takeaway
A four-panelist session on using AI inside a family office landed on a deliberately conservative line — AI's job is to get you to conviction faster and save time, not to make the investment call — and then split openly on the question that matters most: how much of your private data you feed a model in exchange for better output. The spectrum ran from a fully air-gapped private stack to deliberately prioritising velocity over data security, with a sharp counter that an AI fed only your own data is "talking into a tomb" and will eventually have to branch out for better decisions. The practical through-line: treat the model as a discussion partner, not a search box, and expect internally built tools to be overtaken by off-the-shelf ones within months.
Key points
- The panelists spanned very different shops: Melanie, a co-CEO running operations and the investment cycle at a family office she joined this February, from a critical-infrastructure background; Nicholas, who works with Justin Kan (the Twitch founder, sold to Amazon a decade ago) and focuses on physical AI; Fritz, from a 15-person Berlin family office of 12 years; and Andrew Ofori of Bloomvine, who deploys AI to originate private-market deals against investors' search criteria.
- "Garbage in, garbage out" was the opening consensus. Melanie's first initiative (run with Accenture) was building a clean, reliable data structure, because without clean data the output is worthless — paired with training people and putting guardrails on which tools may and may not be used, since a breach is easy and "not good for our families where everything goes out."
- Nicholas framed the build-versus-buy call: screening hundreds of deals a year (he put the ceiling near a thousand), his firm tried building AI tools in-house and concluded it was better to invest in and copilot with a specialist — its seed investment Originalis, which lets him push every deal and all his notes through one platform. His governing line: AI gets you to conviction faster; it does not replace the humans evaluating companies and funds. Proof point — they ran a deal they were going to do past Originalis, which flagged it.
- The human-judgment boundary recurred. Fritz: AI is excellent at extracting data and producing the result you ask for, but once you move past data into judgment-driven output, human discretion is essential. Andrew: the danger is treating AI as a panacea for origination — you still need different kinds of people, and tool use is highly specified, since a commoditised volume search (an HVAC roll-up) and a narrow industrial mandate need very different setups.
- Tool sprawl and pace were a shared frustration: one investor described facing five to seven new vendors every week and wanting a single workflow rather than ten separate tools — AI as a smarter CRM that actually reaches into the underlying datasets. Rogo was named as the standout already widely used in investment banking.
- The most useful workflow details: Melanie runs Claude with named "challenger" and "risk manager" personas to pressure-test a decision paper, then feeds in her team's drafts to ask what is missing and where manual checks are still required. Another approach builds dedicated analyst agents — a VC-fund analyst, a crypto hedge-fund analyst, a crypto-venture analyst, a public-markets and macro analyst — with a feedback loop to refine them over time.
- For AI-assisted diligence, one shop turned a qualitative read into a repeatable screen: judging whether a target had a culture of precision and cleanliness used to mean studying photos, and is now done by criteria — with a blunt elimination rule that anyone with "dirty workspaces" is an automatic no.
- The strongest practical advice: treat the model as a discussion, not a query. Ask Claude or ChatGPT to "be your CTO" and direct you in building; set a six-month goal of spending ten to twenty minutes daily across three or four different LLMs to learn what each does, remembering that context does not carry between them. And the build-versus-buy lesson with a date attached: a deal-sourcing tool one shop built in-house as proprietary IP became pointless "since February," because skills and APIs now replicate it cheaply — so build if you must, but expect to be overtaken.
Notable claims, calls, or numbers
- Deal flow in the hundreds to roughly 1,000 per year; the explicit payoff is time saved getting to information faster on a large share of them, not the decision itself.
- Originalis (seed, backed by Nicholas's firm) flagged a deal the firm was otherwise going to do.
- Broad agreement on a wishlist item: a DDQ-standardisation tool, since building a DDQ from scratch is a slow, bad process for both GPs and LPs.
- An in-house, proprietary deal-sourcing tool was rendered redundant "since February" by off-the-shelf skills and APIs.
- Named tools in use: Claude (run with challenger and risk-manager personas), ChatGPT, Fountain Arc (private full-stack AI), graas.ai and inven.ai (sourcing and aggregation), Jamie (report preparation), Pegasus, and Rogo (investment banking).
- The hard rule, stated flatly: "I will never put a complete list of my family assets into Claude."
Disagreements or tensions
- Privacy versus velocity was the real split, and the panel did not paper over it. At one end, a fully private posture: a complete, air-gapped private AI stack called "the only way," NDAs signed with managers, and an in-house AI server in Sweden running private models — accepting "far inferior" results (a gap that is narrowing) as a deliberate incremental-risk decision. At the other end, Fritz's shop made the opposite call: prioritise velocity over data security, put everything in, and accept the exposure — rationalised by being mostly public-market focused, and rewarded with materially better output when Claude can see the internal portfolio structure and the firm's past decisions.
- The sharpest counter came from the room: an AI fed only your own data is "talking into a cave," a tomb of your own material — fine early on if you have a large volume to process, but self-limiting, because you eventually have to branch out of that tomb to get better investment decisions. Nicholas sat in the pragmatic middle (lean on a specialist partner and tooling, never hand over the full asset list), and several noted the privacy calculus is geographic and stage-dependent: data is "two different conversations" in Europe versus America, and an early-stage investor shrugged that leakage barely matters because the round closes anyway.
Implications for portfolio positioning
- Treat AI as operating leverage on the investment process — sourcing, extraction, drafting, faster conviction, and a unifying workflow over scattered tools — not as a decision-maker; keep a human owning every judgment-driven output.
- Make the privacy-versus-velocity tradeoff an explicit, written policy rather than a default. The panel showed four defensible positions: air-gapped private models at the cost of quality; full-context public models at the cost of exposure; a specialist partner with hard limits on what data leaves; and the "branch out of the tomb" argument that purely internal data caps decision quality. The right mix depends on how public-market versus how sensitive your book is, and on geography and stage.
- Clean data and guardrails are the precondition, not an afterthought, plus training so staff know what may and may not be fed to a model.
- Build-versus-buy now leans hard toward buying: an internally built sourcing tool was obsolete within months, so favour specialist partners (Originalis, Fountain Arc) and task-specific tools (graas.ai, inven.ai, Jamie, Pegasus), with a standardised DDQ workflow the clearest near-term win.
- On adoption, the prescription was hands-on: make the model a "discussion" not a query, ask it to "be your CTO," commit to daily practice across several LLMs, and as a manager push colleagues to build the habit (even hiring a freelancer to seed a specific topic).
- The line worth adopting verbatim: never put a complete list of family assets or other identifying data into a public model — consistent with running sensitive work in controlled environments rather than pasting confidential material into third-party tools.
- The conservative consensus: AI's job is to get you to conviction faster and save time, not to make the call — keep a human owning every judgment-driven output. Nicholas (who works with Twitch's Justin Kan) copilots with seed investment Originalis, which once flagged a deal they were otherwise going to do.
- The real split was privacy versus velocity: a fully air-gapped private stack (an in-house AI server in Sweden, "far inferior" results) versus deliberately prioritising velocity over data security (Fritz puts everything in and gets better output when Claude sees the internal book) — with a sharp counter that an AI fed only your own data is "talking into a tomb" and must branch out for better decisions.
- Clean data and guardrails are the precondition ("garbage in, garbage out"), plus training so staff know what may and may not be fed to a model. The privacy calculus is geographic and stage-dependent — different in Europe versus the US, and early-stage investors shrug at leakage because the round closes anyway.
- Build-versus-buy now leans to buying: an in-house sourcing tool built as proprietary IP was obsolete "since February," so favour specialists and task-specific tools — Originalis, Fountain Arc, graas.ai, inven.ai, Jamie, Pegasus, and Rogo in banking — with a standardised DDQ workflow the clearest near-term win.
- Adoption is hands-on: treat the model as a discussion, not a query, ask it to "be your CTO," practise daily across several LLMs (context does not carry between them), and build dedicated analyst agents with feedback loops. The line worth keeping verbatim: never put a complete list of family assets into a public model.
Preparing for World War Three
Panel discussion · speakers attributed by role
Headline takeaway
Despite the title, the panel's core argument was that World War Three in the conventional sense is not in the cards — no major power has the capability or motivation for global conquest. The real risk to plan for is destabilization: cyber attacks, infrastructure disruption, and foreign interference that trigger societal panic far out of proportion to the physical damage. The investment framing the moderator pushed was to treat "WW3" as a stand-in for unexpected tail risk and ask how you structure a portfolio to survive shocks you cannot predict.
Key points
- The panel was framed through Strauss-Howe generational theory, which describes roughly 25-year secular cycles moving from crisis to rising to crisis again, with the view that we are in the first phase of a new crisis period.
- The moderator offered two ways to take the conversation — literally, through the cycle theory, or as a symbol for unexpected risk and how to structure a portfolio against it — and signalled a lean toward the second.
- The central thesis on why a global war is unlikely: no one today has the means or motivation to fight one. The panel broke modern warfare into three types.
- Type one, legacy conventional warfare, is the Russia-Ukraine model. A panelist cited the Russian advance at roughly 40km over four years — an average of a few metres per day — at a cost of nearly a million men and around 75,000 combat vehicles, making it wildly ineffective for territorial gain relative to the human and material cost.
- Type two, asymmetric warfare, was illustrated by Iran, Azerbaijan, and Iraq (with Vietnam and Afghanistan as historical analogues). It is politically effective — you avoid major battles against a superior enemy, preserve your forces, and use denial and disruption to cause attrition and pressure the other side's political system — and it can secure a draw against a stronger power. But it cannot scale: you cannot invade and occupy a continent with small drones and IEDs.
- Type three, contemporary advanced warfare, is superior in capability but prohibitively expensive. The panelist cited a B-21 bomber at roughly $750M each, an F-22 at roughly $350M each, and autonomy munitions at roughly $2.6M per shot, and noted that a single ~40-day regional war depleted somewhere between 30% and 50% of US munitions. No one has the industrial base to sustain that at large scale.
- On the specific WW3 scenarios, the panel argued the Chinese army lacks proven amphibious-assault capability, that large-scale desert crossings are virtually impossible, and that the industrial base for sustained advanced warfare simply isn't there — so conquest-style global war is not the realistic threat.
- The threat that is realistic is destabilization — foreign interference directed at domestic populations rather than armies in the field, with a Venezuela invasion attempt offered as an example of the kind of destabilizing action to watch.
- On intelligence, the panel argued parity has been reached: the Five Eyes advantage (Australia, Canada, the US, New Zealand, the UK) is now matched by the intelligence apparatus of several adversarial nations, eroding a long-standing Western edge.
- The most vivid point was infrastructure fragility. A panelist argued that if a city like Atlanta had its power grid taken down for even seven or eight minutes, the societal effect would be cataclysmic — hundreds of thousands of people with emergency services down and no ability to call 911 — and that this kind of brief disruption, not a Hollywood long-range nuclear strike with a 90-minute countdown, is what would actually cause panic.
- The summary conclusion: modern conflict won't resemble historical warfare — there will be no Normandy-style operations. Brief infrastructure disruptions create massive societal panic, and the international system for resolving conflict can no longer be relied upon.
Notable claims, calls, or numbers
- Russia-Ukraine: ~40km of advance over four years (a few metres a day on average) at a cost of nearly 1M men and ~75,000 combat vehicles.
- Cost of advanced warfare: B-21 ~$750M each, F-22 ~$350M each, autonomy munitions ~$2.6M per shot; a ~40-day regional war depleted 30-50% of US munitions.
- A 7-8 minute power outage in a major US city was used as the canonical example of how a brief, cheap disruption produces cataclysmic societal effects.
- Strauss-Howe framing: ~25-year cycles, with the claim that we are in the first phase of a new crisis period.
Disagreements or tensions
This was largely a shared-frame discussion rather than an adversarial one; the panelists were aligned that conventional global war is unlikely and that destabilization is the real risk. The productive tension was methodological — whether to engage the generational-cycle theory on its own terms or to treat the whole topic as a proxy for unpredictable tail risk and focus on portfolio resilience. The panel leaned to the latter.
Implications for portfolio positioning
- The actionable takeaway is to structure for unpredictable tail risk rather than to forecast a specific conflict: build resilience and redundancy on the assumption that the disruptive event will be something brief, infrastructural, and panic-inducing rather than a predictable military escalation.
- Treat critical-infrastructure fragility (power, emergency services, communications) and cyber/influence operations as the live exposures, which argues for resilience-oriented holdings and caution on assets acutely sensitive to short, sharp societal shocks.
- The panel's framing implies favouring portfolio construction that can absorb a shock you didn't see coming over precise geopolitical bets — consistent with the "be specific and build in caution" message that ran through the conference's other geopolitics sessions.
Speaker names were not captured for this session; contributions are attributed by role.
- Despite the title, the panel's core argument is that conventional World War Three is not in the cards — no major power has the means or motivation for global conquest. The moderator pushed reading "WW3" as a stand-in for unpredictable tail risk.
- Modern warfare breaks into three self-limiting types: legacy conventional (Russia advanced ~40km in four years at a cost of nearly 1M men and ~75,000 vehicles), asymmetric (politically effective but cannot occupy a continent), and advanced (prohibitively expensive — a B-21 at ~$750M, an F-22 at ~$350M, autonomy munitions at ~$2.6M a shot, with a ~40-day war depleting 30-50% of US munitions).
- The realistic threat is destabilization — cyber attacks, infrastructure disruption, and foreign interference aimed at populations rather than armies — and the long-standing Five Eyes intelligence edge has eroded toward parity with several adversaries.
- The most vivid risk is infrastructure fragility: taking a major city's power grid down for even 7-8 minutes (no 911) would cause societal panic far out of proportion to the physical damage — brief and cheap, not a Hollywood nuclear countdown.
- The portfolio implication is to structure for shocks you cannot predict rather than forecast a specific conflict: build resilience and redundancy, and be cautious on assets acutely sensitive to short, sharp societal shocks and on critical-infrastructure and cyber exposure. Not an investment session, but the framing carried across the conference's geopolitics sessions.
Crypto 2027: Industry Update and Outlook
Matt Hogan (Chief Investment Officer, Bitwise)
Headline takeaway
Bitwise's Matt Hogan led a panel that reframed crypto as three separate conversations — bitcoin as a monetary asset, stablecoins as financial rails, and a new generation of revenue-generating tokens — and argued the asset class is now "a growth asset trading at value levels" because retail selling has masked relentless institutional buying. The most concrete calls: a long-run path to $1M bitcoin if it captures 20% of gold's store-of-value market, stablecoins settling on a handful of public blockchains, and revenue tokens like Hyperliquid that should be judged on revenue multiples rather than P/E.
Key points
- Setting the scene, Matt Hogan described a tale of two stories: enormous regulatory progress and billions of bitcoin bought by institutions and sovereign wealth funds, yet a falling price. The reconciliation is that there are two kinds of crypto investor — institutions are marginal new buyers, but retail still controls about 70% of the market and has sold more than the institutions bought. Retail selling has slowed while institutional buying has continued, so the two groups are moving in opposite directions.
- Hogan argued crypto has become three distinct conversations: bitcoin as a new global monetary asset and hedge against fiat; stablecoins and fintech as a legitimate new set of payment rails, which resonates most with the traditional financial community; and revenue-generating tokens such as Hyperliquid trading at attractive multiples. He flagged an anchoring bias — most people are not aware of this newer, revenue-driven generation of crypto.
- A panelist (John Silver) summarised the setup as crypto being "a value stock" — a growth asset trading at value levels — and added that it is actually positive that AI has taken the oxygen out of the room, alongside genuine progress in institutional adoption.
- Another panelist (Sergey) noted that the transaction volume of USDT has exceeded Visa and Mastercard, and that while a P/E multiple is meaningless for these assets, a revenue multiple and value-capture analysis still make sense — they should not be treated as traditional stocks.
- On blockchain as fintech rails, the panel used the "internet before broadband" analogy: five years ago the ecosystem had low throughput and variable transaction costs, making real financial rails impossible; now blockchains can handle around 100,000 transactions per second at low fees. Two examples grounded it: stablecoins now account for around 10% of remittances to Mexico (the second-largest channel), at an on-ramp-to-off-ramp dollar-to-peso cost of roughly 2.5% versus 6-8% on standard forex, and DoorDash has rolled out paying its dashers in stablecoins.
- The second example was Hyperliquid, a perpetual-futures exchange whose design traces to a Nobel-prize-winning concept, operationalised in crypto. It runs 24/7/365, is doing a meaningful chunk of oil-derivatives volume with about half its activity coming from non-crypto applications, generates roughly $1bn of revenue a year with 100% of it going to buy back the token, and is now used by macro hedge funds when news events hit.
- On regulation, the panel separated the Clarity Act from the Genius Act (stablecoins), and argued the real impact will not be felt until around June 2027. The expectation is that all major Wall Street firms will roughly 10x their blockchain investment, alongside a VC boom of billion-dollar funds (a16z and Haun were named raising), with Canton, Solana, and Ethereum among the likely beneficiaries.
- Asked whether stablecoins will run on public blockchains, the panel said yes — "open wins," per the internet analogy — but expects an oligopoly with some regionalism (Tron already dominates Asia). One panelist named six likely winners: Ethereum, Solana, Canton, Arc, Tempo, and possibly Avalanche, and expects the space to 10x even though no one knows exactly how.
- On dead use cases: a panelist who had built a crypto-gaming startup described the 2-years-ago hype around the Telegram and TON ecosystem, which ultimately collapsed to the question of whether crypto was even needed and why a new blockchain was required. On DeFi, the view was more nuanced — despite hacks there are still opportunities, it is not as hot as a year ago but not in full winter, most tokens are severely depressed, and only a few projects with robust value-capture mechanisms (BTC, ETH, SOL, and a public-company lens, with YieldBasis named) are worth holding.
- On token longevity, the panel's rule of thumb: a token continues to exist as long as at least one person maintains the protocol.
- On the bull case for bitcoin, Hogan laid out the $1M thesis: bitcoin is competing with gold, and as Kevin Warsh put it, bitcoin is "gold for people under 40." With gold around a $30trn market and growing roughly 12% a year for 20 years, bitcoin only has to take about 20% of the store-of-value market over ten years to reach $1M per coin. The catalysts: the passage of time, the unlocking of new platforms and institutions, and rising concern about US debt and deficits feeding a debasement trade.
Notable claims, calls, or numbers
- Retail still controls ~70% of the crypto market and sold more than institutions bought over the past year; institutional buying has continued throughout.
- Stablecoins are ~10% of remittances to Mexico at ~2.5% all-in cost versus 6-8% on forex; USDT volume has surpassed Visa and Mastercard.
- Hyperliquid: ~$1bn annual revenue, 100% directed to token buybacks, ~half of activity non-crypto, used by macro hedge funds on news events.
- Genius Act impact not expected until ~June 2027; major Wall Street firms expected to ~10x blockchain investment.
- Bitcoin $1M thesis: gold ~$30trn market growing ~12%/yr; bitcoin needs ~20% of the store-of-value market over ten years.
- Bitcoin ETF flows: long-only investors' share has risen from 5% to 40% over two years and is expected to settle around 60%; ETFs only recently approved on Morgan Stanley and Wells Fargo platforms, with first flows in Q1; global equities ~$10trn versus bitcoin ~$2trn implies a ~2% starting point.
- The gold-ETF analogy: launched 2004, $3bn in year one, six consecutive years of inflows — adoption takes time.
- Bitwise's tokenised basis-trade fund (USCC) was cited at a 5.1% return in USD terms, with basis trades also referenced in SOL and XRP.
- MicroStrategy holds a ~$60bn bitcoin position with relatively little debt and obligations covered for years.
Disagreements or tensions
- The panel split, productively, on why to own bitcoin at all. The bull case rested on the monetary-asset thesis and ETF adoption curve; the skeptical case, raised in the room, was that you diversify to capture a better risk-adjusted return profile, and bitcoin has been more correlated with equities than its "digital gold" billing implies, which weakens the diversification incentive — especially with two years of ETF inflows already in the price.
- On risks, quantum computing was the live debate: it is already slowing institutional appetite and causing roughly six-month delays while the industry explains how bitcoin copes. The specific exposure is old wallets whose public keys are revealed, including an estimated 3-4M BTC in Satoshi's wallets that, if Satoshi is dead, cannot be moved to a quantum-protected wallet. Solutions exist (reserving that BTC in shared custody and requiring direct claims), but the panel acknowledged the hard part is getting a decentralised community to agree — while insisting bitcoin "won't walk off a cliff."
- A second risk flagged was headline/regulatory shock — a sudden anti-bitcoin regulatory response, with an Iran-related episode cited as the kind of event that can jolt the price.
Implications for portfolio positioning
- The cleanest framing for an allocator is to treat crypto as three different allocations, not one: a monetary-asset position (bitcoin), an infrastructure/rails position (stablecoins and the chains that win), and a revenue-token position judged on revenue multiples and value capture rather than P/E.
- The institutional adoption curve is the structural tailwind: with long-only ownership of ETFs rising from 5% toward a projected 60% and a ~2% starting weight versus global equities, the flow argument is that inflows continue for years, mirroring the gold ETF's slow ramp.
- For yield, the tokenised basis trade (Bitwise's USCC at ~5.1%) is a concrete, relatively understandable income strategy versus directional exposure.
- Weigh the two genuine risks the panel named — quantum (a multi-year overhang, not an imminent break) and headline/regulatory shocks — against the debasement-trade thesis tied to US debt and deficits, which the panel saw as the main long-run catalyst.
- On structure risk, the panel distinguished MicroStrategy (a unique, low-debt position) from its smaller, more leveraged imitators, which it saw as the more fragile actors though not a systemic threat.
- Treat crypto as three separate allocations, not one (Bitwise's Matt Hogan): bitcoin as a monetary asset, stablecoins as payment rails, and revenue-generating tokens judged on revenue multiples rather than P/E.
- The setup is "a growth asset trading at value levels": institutions are buying relentlessly while retail — still about 70% of the market — has sold more than they bought, so the two are moving in opposite directions and price has lagged adoption.
- The bitcoin $1M thesis: gold is a ~$30trn market growing ~12% a year, and bitcoin only has to take about 20% of the store-of-value market over ten years, helped by a debasement trade tied to US debt and deficits. ETF long-only ownership has risen from 5% to 40% and is expected to settle near 60%.
- Rails are now real (~100k transactions/sec): stablecoins are ~10% of remittances to Mexico at ~2.5% all-in versus 6-8% on FX, USDT volume has passed Visa and Mastercard, and Hyperliquid generates ~$1bn of revenue with 100% directed to buybacks. For yield, Bitwise's tokenised basis-trade fund was cited around 5.1%.
- The genuine risks were named, not averaged away: quantum as a multi-year overhang (old wallets, ~3-4M BTC in Satoshi's wallets) and headline/regulatory shocks, with the Genius Act's impact expected around June 2027. The bear case from the room: bitcoin has tracked equities more than its "digital gold" billing implies, weakening the diversification reason to own it — and the panel distinguished MicroStrategy's unique low-debt ~$60bn position from its more leveraged imitators.
A Snapshot of Longevity Medicine: What Works in the Clinic Today and Where Will We Be Tomorrow?
David Sinclair (Geneticist, Harvard; Founder, Life Biosciences) · Todd MacAllister (Longevity clinics (Therme Group))
Headline takeaway
A longevity panel split cleanly between near-term "wellcare" delivery and frontier "reset the clock" biology, and the most useful investor takeaway was a diligence rule, not a product: Harvard's David Sinclair urged allocators to go to the researchers, not the people monetising the headlines, because the field is full of conflicts of interest. The substance underneath the hype: epigenetic reprogramming has moved from yeast and mice into human eye trials, and the cleanest validation signal is that big pharma is now doubling down and bringing AI drug discovery in-house.
> Note: the live recording captured only a fragment; this summary is reconstructed from contemporaneous notes taken in the room.
Key points
- David Sinclair (Harvard, 25 researchers, 25 years; founder of Life Biosciences) framed the honest version of the field: short term there is real hype — social-media pundits and even doctors overstating results, often with undisclosed conflicts — but long term the technology is being underestimated. He drew the explicit parallel to AI: we overestimate in the short run and underestimate in the long run.
- The science he is betting on is epigenetic reprogramming. His frame: DNA is the hardware of biology and the epigenome is the software; aging is loss of information, and there is a backup copy of that software in the cell that can be restored. Using a subset of the Yamanaka factors (the Nobel-winning stem-cell genes), his work takes a cell's age back roughly 75%, has rejuvenated monkey eyes, and is now entering human trials.
- Life Biosciences' first human trials target glaucoma and NAION ("a stroke in the eye" — vision lost overnight, with nothing doctors can currently do). Sinclair argued no other company is as far along in humans, and credited an early start plus the core IP. He acknowledged well-funded competition backed by Jeff Bezos and Sam Altman (the Altos Labs cohort).
- The other half of the panel, Todd MacAllister, works the delivery problem: shifting healthcare from reactive "sickcare" to proactive "wellcare." His blunt framing is that until therapeutics actually extend life, roughly 95% of longevity practice is just not dying early — good habits, testing, and monitoring.
- MacAllister's hard problem is not the science but adherence. Through the Therme Group (a wellness-resort operator he cited at around 6 million customers), epigenetic testing and monitoring are the easy part; the difficulty is compliance, because clients typically engage for about 6 months and then quit. The commercial question he posed is how to keep people past that drop-off.
- On environment, Sinclair noted the epigenome is shaped by exposure — nonstick chemicals, what you breathe, the gut microbiome — so people age faster or slower depending on toxic burden, and removing inputs can slow the clock. He claimed a blood sample can already read how fast you are aging.
- On cost and access, the panel traced a steep democratisation curve: longevity programmes that cost about $15,000 per head when MacAllister started are now sold as roughly $25,000-a-year memberships, with a $10,000-a-year tier built last year; Function Health has pushed a testing entry point down to around $400, and a Therme visit runs roughly €30-40.
- On the investment landscape, the panel described the molecule-iteration loop (chemist to mouse and back) compressing from years to months as Silicon Valley AI-bio companies take the chemist out of the early loop. Sinclair called the AI-and-biology union one of the biggest achievements of the century and named the real-world test of seriousness: big pharma doubling down, bringing talent in-house, and using AI for its own drug development. His marquee data point — the former head of GSK now runs Altos Labs.
Notable claims, calls, or numbers
- Partial reprogramming via Yamanaka factors takes cell age back roughly 75%; monkey-eye rejuvenation done, human eye trials underway.
- Life Biosciences' in-human trials target glaucoma and NAION; the panel's claim is that no competitor is as advanced clinically.
- An 8-point improvement on the relevant health marker was cited as cutting mortality by about 40% and adding roughly 4 years of health expectancy.
- Cost curve: about $15,000 per head historically, now ~$25,000/year memberships and a $10,000/year tier; Function Health ~$400 entry; Therme ~€30-40 per visit; Therme cited at ~6 million customers.
- A 2020 Nature cover paper marked the shift from slowing aging to resetting tissue age.
- The molecule-iteration loop is compressing from years to months as AI front-loads the chemistry.
- Signal of capital and talent into the field: the former head of GSK now leads Altos Labs; Bezos and Altman are backing competitors.
Disagreements or tensions
- The productive split was between the two business models on the stage. MacAllister's "wellcare" is behavioural — testing, habits, and a real retention problem — and delivers value today; Sinclair's reprogramming is therapeutic and aims to restore tissue function modern medicine cannot, but is still in trials. They are complementary on his own timeline ("in 4 years take the wellness work, in 10 years take the epigenetics work"), but they are different risk and time horizons for an investor.
- The sharper tension is hype versus substance. Sinclair was candid that the short-term noise is real and conflicted, which is precisely why he pushed diligence toward primary researchers and big-pharma behaviour rather than the social-media layer.
Implications for portfolio positioning
- Treat longevity as two distinct allocations, not one theme: a near-term consumer-wellness business (clinics, testing, resorts) that lives or dies on retention economics given the ~6-month churn, and a long-duration therapeutics bet (epigenetic reprogramming) priced like clinical-stage biotech.
- Use the panel's own validation tests as a screen: primary research and IP depth over marketing, and big pharma in-housing AI drug discovery as the signal that a sub-field is real.
- The AI-bio discovery-loop compression (years to months) is the structural tailwind; Sinclair's framing implies a few large winners, so back people, IP, and track record rather than the narrative.
- Watch Life Biosciences' glaucoma and NAION readouts as the near-term proof point, and note the competitive field is already well capitalised (Bezos, Altman/Altos).
- Two distinct businesses shared the stage: near-term "wellcare" (clinics, testing, habits) that delivers value today, and frontier epigenetic reprogramming that aims to reset tissue age but is still in trials — different risk and time horizons for an investor.
- The diligence rule, from Harvard's David Sinclair (founder of Life Biosciences): go to the researchers, not the people monetising the headlines. Short-term hype is real and conflicted; long-term the technology is underestimated, and he framed the epigenome as biology's software with a restorable backup copy.
- The science has moved from yeast and mice into human eye trials (glaucoma and NAION): partial Yamanaka-factor reprogramming takes a cell's age back roughly 75%, and Sinclair claims no competitor is as far along clinically, against well-funded rivals backed by Bezos and Altman (Altos Labs).
- The clinic layer (Todd MacAllister, with Therme at ~6 million customers) lives or dies on retention — clients churn at about 6 months — along a steep pricing curve: roughly $15k per head historically, now ~$25k/year memberships and a $10k tier, with Function Health near $400 and a Therme visit ~€30-40. Environment also shapes the epigenome (toxic burden, gut microbiome), and a blood sample can already read how fast you are aging.
- The cleanest validation signals are big pharma in-housing AI drug discovery (the former head of GSK now runs Altos Labs) and the discovery loop compressing from years to months — Sinclair called the AI-and-biology union one of the century's biggest achievements — so back people, IP, and track record, and watch Life Biosciences' eye-trial readouts as the near-term proof point.
The Fracturing World Order: Geopolitical Risks, Markets, and Strategic Capital
Stephen Harper (Former Prime Minister of Canada) · Sir William Browder (Founder, Hermitage Capital Management) · Sir Nick Carter (Former Chief of the Defence Staff, UK)
Headline takeaway
A panel of heavyweight statesmen — former Canadian PM Stephen Harper, Hermitage's Sir William Browder, and former UK Chief of Defence Staff Sir Nick Carter — agreed by show of hands that the US is now the single largest contributor to global instability, not because it is weak but because it has decided that acting as the architect of the global system is no longer in its interest. The investment conclusion they converged on: the US still deserves a core allocation as the deepest, most dynamic market, but the premium it has historically commanded is eroding, so the prudent move is to diversify the marginal dollar away from it while building security and resilience into the portfolio itself.
Key points
- The framing question was how to invest when the architect of the global order moves from being "the house" to being "a player." The panel accepted that premise as already true.
- Stephen Harper argued the US shift is structural, not just a Trump phenomenon: the general direction is that being a systemic stabiliser is no longer seen as in America's interest, and the world is moving toward big-power politics. He agreed the US is the greatest current source of uncertainty (tariff wars, Middle East conflict) but stressed it is the country hurt least, because it is an internally driven economy far less geopolitically exposed than others.
- Sir William Browder put the raw military blame on Russia — Georgia, Moldova, Crimea, Ukraine, and, in his view, the Baltics and Poland next. But his larger point was that despite the US running a tariff war, an Iran war, cutting Ukraine military aid, and threatening NATO obligations, the world has continued to survive with no economic cataclysm.
- Browder's striking claim: Ukraine losing US military aid has been a blessing, because Ukraine is now hitting Russian oil refineries with no US interference. The US is not the military shield it was assumed to be; superpowers are no longer superpowers, and small powers are defending themselves fine.
- Sir Nick Carter focused on reliability: Trump's unpredictability disrupts America's natural allies, and the world now has to manage without depending on US reliability. He argued US big tech introduced populism, which has changed the character of politics and, in turn, the character of warfare — and that there is no international system to set boundaries or guardrails on technology, with the Trump-Xi axis doing nothing about it.
- A recurring lens from Carter: history should be read through technology, not kings and queens — leaders are "musicians playing the instruments of technology." New asymmetric, distributed capabilities equalise power between small and large countries, which introduces more violence and uncertainty.
- Carter noted Sudan is now the most costly war in human casualties, surpassing Ukraine, in a world where "right is might," the West is less able to police it, the UN is no longer the institution that matters, and the established norms for when force is legitimately applied have broken down.
- Browder's technology case study: Russia had an enormous navy operating from Sevastopol and Crimea, and Putin prized Crimea partly for that Black Sea fleet. Ukraine, with no navy at all, used sea drones to neutralise it, so Russia can no longer control the Black Sea. He extended the point to sanctions — he has spent years urging Western governments to sanction the Russian oil industry, but everyone nods and no one touches the oil, so Ukraine effectively imposed its own sanctions by bombing refineries. His theme: technology now lets actors bypass the formal global order to impose outcomes themselves.
- Harper's caution on the technology-equalisation thesis: it is real but should not be exaggerated. Middle powers must operate across two technology stacks, the American and the Chinese, with no viable alternative. From a Canadian vantage point, the play is to build a middle-power consensus around continuing rules while admitting honestly that, especially technologically, they remain inside the American orbit.
- On Europe as a potential power: the panel debated whether Trump is, perversely, the best thing to happen to Europe — pushing it to change its tech stack, challenge Visa and Mastercard, and build domestic defence — and whether that leads to federalised tax, armies, and eurobond issuance. Harper was skeptical: the EU is a unique but extremely slow-moving construct, always a work in progress, and its major states are as reluctant as anyone to transfer power to smaller members. His sharper point: Trump has exposed how economically and militarily dependent on the US the West became, and the real question is whether the will to change that outlives his presidency — historically, political will dissipates the moment the pressure does, and that dependence has made allies underachievers.
- Browder, just back from the Munich Security Conference, noted Vance had chastised the EU as hostile to free speech. He countered that Europe's population is larger than America's, its GDP is 60-70% of the US, and its aggregated military personnel is large; Germany is spending €600bn on defence and Poland 5% of GDP. Europe is not swashbuckling, but in genuine crises — he cited Covid — it coordinates coherent policy, and in a mass-casualty event it will set aside differences and do what is necessary.
- Carter explained Europe's strategic incoherence: if you border Russia, Russia is the problem; if you are in the south, it is Africa and terrorism. Those bifurcated priorities, plus terrible military interoperability, mean Europe needs a shock to unite around a common cause.
- On rebuilding trust: Harper said it takes far longer to build trust than to dismantle it, and predicted a future Democratic president will still be protectionist — there is no return to the globalised, multilateral, market-rules security approach. Browder was more optimistic: people blame Trump personally rather than America, and in roughly 2.5 years a new president will go on a "resetting relations" tour that every head of state will gladly welcome, much as every new US president has reset with Putin. Carter argued the institutions, not the president, are what hold the system together — the UK's "special relationship" rests on intelligence sharing, Five Eyes, 45,000 troops, and complete system access, and the same institutional ties anchor Canada, New Zealand, and Australia.
Notable claims, calls, or numbers
- Germany is spending roughly €600bn on defence; Poland is spending 5% of GDP.
- Europe's GDP was put at 60-70% of the US, with a larger population.
- A panelist's own portfolio was historically 80% US / 20% rest-of-world, and he is now questioning whether the next marginal dollar should go to the US at all — not because the rest of the world is better, but because the US premium looks unsustainable.
- India was cited as the 4th largest economy by the end of this year, with some suggesting 2nd largest by the end of the decade.
- The clearest portfolio prescription: treat security as an investment variable — IT, communications, trade, and health-supply-chain risk — by shortening supply lines, building redundancy, and being cautious about cross-border exposure that can be disrupted by geopolitical events.
Disagreements or tensions
- The sharpest divergence was on US exceptionalism. Harper insisted you should diversify and add caution but not turn away from the US, which remains the largest, most dynamic, entrepreneurial economy and is relatively insulated as an energy exporter. Browder agreed the US has the deepest capital markets and reserve-currency advantage — noting the UK would already be in a debt crisis if it ran America's deficit without a reserve currency — but argued the age of US exceptionalism is ending and the premium justifies moving money away.
- On near-term macro, Harper flagged a serious risk of stagflation for developed economies (though he said we are not truly there yet) and warned that broad macro strategies are the wrong approach — investors need to be very specific to individual markets. Browder framed the tension differently: military spending is rising and oil prices and taxes with it, but a massive technology shift could change the cost of everything, leaving many investors paralysed between the geopolitical fears and the AI boom, and therefore holding tight.
- On China and Taiwan, Carter pushed against the consensus hype: he does not think China needs to invade, because a classic Sun Tzu political-warfare strategy will subsume Taiwan over time — cutting internet cables, bringing adjacent regions under Chinese telecom influence, seducing the opposition, and benefiting from younger Taiwanese warming to China partly in reaction to Trump.
Implications for portfolio positioning
- Keep a core US allocation but stop treating it as the automatic home for the marginal dollar; the historical premium is eroding and diversification is now the prudent default rather than a hedge.
- Build security and resilience into the portfolio as first-order variables: shortened supply chains, redundancy, and caution on assets exposed to border or infrastructure disruption.
- Tilt toward resilient, energy-self-sufficient, institutionally stable economies — Canada and Australia were named — and take seriously the long-horizon growth case for India and, more speculatively, Africa.
- Prefer specificity over broad macro bets given stagflation risk; the panel was explicit that blunt top-down strategies are poorly suited to this environment.
- On the UK specifically, the bull case is soft power and London's enduring role as the centre of international business, even through domestic political turbulence; the bear case (Harper) is post-Brexit governance with no positive change in sight.
- The panel agreed by show of hands that the US is now the single largest source of global instability — not from weakness but a deliberate choice that being the system's stabiliser no longer serves its interest. Harper called the shift structural and bigger than Trump: the world is reverting to big-power politics.
- Investment conclusion: keep a core US allocation (deepest market, reserve-currency and energy-exporter advantages) but stop sending it the marginal dollar — the US premium is eroding, so diversification is now the default rather than a hedge. One panelist's 80% US / 20% rest-of-world book is under review.
- Technology, not armies, now equalises power: Ukraine, with no navy, used sea drones to break Russia's Black Sea fleet and bombed refineries to impose the oil sanctions the West wouldn't — actors increasingly bypass the formal order to force outcomes themselves.
- Tilt toward resilient, energy-self-sufficient, institutionally stable economies (Canada and Australia named) and take India's long-horizon growth seriously (4th-largest economy this year, possibly 2nd by 2030). Browder put Europe's GDP at 60-70% of the US, with Germany spending €600bn on defence and Poland 5% of GDP.
- Prefer specificity over broad macro bets given stagflation risk, and treat security — supply chains, redundancy, infrastructure exposure — as a first-order portfolio variable. The disagreement worth keeping: Harper says diversify but don't turn away from the US; Browder says the age of US exceptionalism is ending and the premium justifies moving money out.
A Changed World Order: Its Impact on Military Spending and Modern Battlefield Technologies
Klaus Hommels (Founder and Chairman, Lakestar)
Headline takeaway
A defence-investing panel — anchored by Lakestar founder Klaus Hommels alongside an executive from the drone manufacturer Tekever — made the case that defence has become investable for family offices for the first time, because patriotism and economic return now point the same way. The practical advice: the binding constraint is not capital but red tape and procurement, the best venture entry is backing startups with strong M&A prospects into the primes, and the durable thesis is counter-technology that defeats today's cheap drones and expensive missiles.
> Note: this panel was recorded in two parts. Only the first part's notes are recoverable — the second recording captured no audio. The summary below reflects the available material.
Key points
- An opening speaker placed the discussion in the context of new warfare dimensions — space and biology, with China reportedly calling biology the "sixth dimension of warfare" — and warned investors to be wary of the expert opinion they can access: when cyber first became an investable theme, around 90% of the firms selling cyber software were effectively scams, and the same dynamic is now playing out in defence.
- Klaus Hommels framed the core questions: what is the space for nimble suppliers and early-stage startups, and how do public contractors and private markets interact? He noted that around 90% of new technology companies are coming out of Germany and the UK.
- A structural problem he flagged: defence is not like B2C, where you can sell to the world. A defence company's first contract has to come from its own national military, and small countries cannot offer contracts large enough to anchor a company — even though Europe still needs to buy the same equipment.
- His investment prescription: given the sales cycles and the importance of partnerships, the best way to invest is to pick startups with strong M&A prospects, because they will ultimately be taken over by the primes.
- On why families have historically avoided defence, Hommels argued you have never before had the combination of patriotism and economic outlook aligned — and that the real risk and source of complacency today is red tape, citing 650 documents required to apply for a licence and a system very slow to adapt. His historical aside: in the first two months of WWII, the UK spent more on horses than on gasoline.
- For how a family office actually invests in defence, he pointed to the economics of the battlefield — missiles costing around $1M versus drones costing around $15k — and argued the opportunity is in counter-technologies: any company that can deny the capabilities of current missiles and drones.
- The Tekever executive introduced the company as a leading surveillance-drone manufacturer that had zero defence sales five years ago and is now building the largest drone manufacturing operation across several NATO countries — describing it as Europe's only profitable defence "unicorn" or neoprime, with 13,000 people across Europe and the US.
- Tekever's mission, after four years operating in Ukraine (present since the first weeks of the conflict), is to bring that learning into NATO countries. The distinction it drew: it builds high-end drones defined by their capabilities, not the simple buy-and-fly hardware most people picture, and what works in a live war like Ukraine differs from what works in the UK, which is not at war — which changes procurement and regulation.
- On economics, the executive cited roughly 70% gross margins and 40% EBITDA margins, and stressed that Tekever was bootstrapped for a long time. Living off customers' money rather than investors' money created management discipline, allowed the company to react quickly, and built a culture centred on developing people.
- On scaling, the executive traced the path: Tekever started in the UK in November 2018 surveying and keeping the Channel safe, and crucially did not sell drones — it operated drones as a service, building a feedback loop in which the people flying the systems are its own employees in the civilian world. It never expected to be in defence until Ukraine. By 2026, Ukraine is a small percentage of revenue; the past two years have been spent working with governments on how they build and sustain capability. The closing point: technology is the key, and the specific product is irrelevant.
Notable claims, calls, or numbers
- ~90% of new defence-relevant technology companies are coming out of Germany and the UK.
- 650 documents required to apply for a defence licence — cited as the real constraint, not capital.
- Battlefield economics: missiles ~$1M each versus drones ~$15k each, which is the basis for the counter-technology investment thesis.
- Tekever: zero defence sales five years ago to Europe's only profitable defence neoprime; 13,000 employees; ~70% gross margin, ~40% EBITDA margin; bootstrapped; Ukraine now a small share of revenue.
- A panelist's framing that defence is investable now specifically because patriotism and economic return are aligned for the first time.
Disagreements or tensions
No open disagreement surfaced in the available notes. The implicit tension for an allocator is between the appeal of the theme and the warning that the sector is full of low-quality vendors riding the wave — the explicit cyber analogy (90% scams early on) is a caution to apply hard diligence rather than buy the narrative.
Implications for portfolio positioning
- Defence venture is best accessed by backing companies with credible M&A exits into the primes, given the procurement and contract-anchoring constraints on standalone scale.
- The most durable sub-thesis named was counter-technology — systems that defeat the current generation of cheap drones and expensive missiles — rather than the drones or missiles themselves.
- Treat the "patriotism plus economics align" framing as the reason the category is now open to family offices, but pair it with the panel's own warning that diligence quality matters more here than in most themes because of the volume of weak vendors.
- Margin and discipline signals matter: the standout operator on the panel was bootstrapped to high margins, which the executive tied directly to reacting quickly and building real capability — a useful screen when evaluating defence-tech managers and direct deals.
- Defence has become investable for family offices for the first time, because patriotism and economic return now point the same way — the panel's anchoring argument, from Lakestar's Klaus Hommels, who noted roughly 90% of new defence-relevant technology companies are coming out of Germany and the UK.
- The binding constraint is not capital but red tape and procurement: Hommels cited 650 documents required to apply for a licence and a system very slow to adapt (his historical aside — in the first two months of WWII, the UK spent more on horses than on gasoline).
- The best venture entry is to back startups with strong M&A prospects into the primes, because a defence company's first contract has to come from its own national military and small countries cannot offer contracts large enough to anchor a standalone business.
- The durable thesis is counter-technology — systems that defeat today's cheap drones (~$15k) and expensive missiles (~$1M) — rather than the drones or missiles themselves.
- Tekever was the operator case study: zero defence sales five years ago to Europe's only profitable defence "neoprime," 13,000 staff, ~70% gross and ~40% EBITDA margins, bootstrapped. It sells drones-as-a-service (its own people fly the systems, creating a feedback loop), has been in Ukraine since the first weeks, and now earns only a small share of revenue there. The standing caution: the sector is full of low-quality vendors riding the wave — when cyber first became investable, ~90% of sellers were effectively scams — so diligence quality matters more here than in most themes. (Recorded in two parts; only the first part's notes survived.)
Stewardship, Spirituality, and Family Wealth: Multigenerational Perspectives
Panel discussion · speakers attributed by role
Headline takeaway
A small, self-selected session on stewarding not just family wealth but a family's sense of purpose. The reframe that ran through it: treat capital as "stored energy" to be stewarded across generations rather than an asset that must compound at all costs, and build purpose into the next generation early instead of retrofitting it later. For an allocator, the most practical thread was that values- and faith-aligned giving is a more natural on-ramp than values-aligned investing — start with the philanthropic portfolio, where no financial return is expected.
> Note: this write-up captures the panelists' contributions; the moderator's remarks are not included. Personal and identifying detail has been kept out by design.
Key points
- The panelists ran established multigenerational family businesses alongside their own separate ventures, and the opening tension was working for the family versus pursuing your own path. One panelist made the case for clean separation — he and his brother exited the family business to pursue their own direction, and framed staying close to the family while being "spiritually separate" as its own kind of success, adding that even if the new venture generated little wealth he would still count it a success because of that alignment.
- Another, third-generation in a family business, described living "two lives" — the core enterprise versus his own technology investing and values-driven outside work — and how the tension with his father eased once the outside work began to complement the family enterprise rather than compete with it.
- The most reused idea was the Kabbalah concept of the "bread of shame": you can survive on it, but a life built entirely of it leaves you wishing you had found more nutrient-dense purpose. The application — most of the family-office world confronts purpose retroactively (join the foundation, sign checks, then look for meaning), so the better path is to build purpose in from the start, including employing children young so they see the power of compounding over decades.
- A fourth-generation participant anchored the discipline thread: a sustained daily meditation practice, silent retreats taken with the whole family, and a 34-year family foundation used as the first training ground for the next generation. His frame — a business can become a "prison" each generation has to break out of, and that act of breaking out is what creates room for the next generation's freedom.
- On faith and the family office, the practical conclusion was that values- or faith-aligned philanthropy is a softer landing than faith-aligned investing: it sidesteps the "is there a faith premium?" debate because no financial return is expected, and even a largely passive donor-advised fund is being allocated like any portfolio, so the philanthropic capital itself can be aligned with the family's values.
- A stewardship-of-time thread distinguished Chronos (clock time) from Kairos (meaningful moments), with meditation framed as "listening" where prayer is "speaking," and a seven-generation outlook (borrowed from an indigenous wealth-transfer case study) offered as the long-horizon frame for decisions.
- On giving culture, the panel contrasted Middle Eastern norms — where giving is sacred and never named publicly — with Asian families' comfort putting a name on the building, and pointed to Maimonides' seven levels of giving as a useful framework. A recurring caution: founders often leave only the instruction to "sell the company and give," with no criteria, so later generations inherit the unenviable task of defining what the giving is for.
Notable claims, calls, or numbers
- The "stored energy" reframe: treat surplus capital as energy to steward across generations, not an asset that must always grow.
- A 46th-generation Japanese inn, run for roughly 1,300 years, cited as the case study for stewardship over growth — chapters of the lineage were wealthy and chapters were not, but the constant was care for others and restoration.
- The "bread of shame" (Kabbalah) as the argument for building purpose into the next generation early rather than retrofitting it.
- Maimonides' seven levels of giving offered as a structured framework for a family's philanthropic strategy.
- The "is it true / is it kind / is it necessary?" three-part filter, as a discipline for saying less and weighting what you do say.
Disagreements or tensions
- Faith in the investment book split the room. The skeptical view doubted any "faith premium" in returns; the conviction view held that values-aligned managers and giving produce better outcomes. The panel's de-risking move was to start with philanthropy, where return expectations are absent, rather than the investment portfolio.
- Separation versus integration ran underneath the whole discussion: the case for cleanly exiting the family business to find personal alignment, set against the more common instinct to fuse personal purpose with the family enterprise — captured in the framing of the business itself as a "prison" each generation must break out of.
Implications for portfolio positioning
- Not an investment session in the usual sense, but two transferable ideas for a family office: reframe surplus capital as stored energy to steward, which reorders decisions toward longevity and purpose rather than pure compounding; and introduce values- or faith-alignment through the philanthropic or donor-advised portfolio first, where no financial return is expected, before letting it touch the investment book.
- The governance lesson on giving: don't hand the next generation only a "sell and give" instruction. Set criteria, use a framework like Maimonides' levels, and treat the family foundation as a deliberate training ground for the next generation rather than a place to park them.
- The reframe that ran through the session: treat capital as "stored energy" to steward across generations, not an asset that must compound at all costs — and build purpose into the next generation early rather than retrofitting it, captured in the Kabbalah idea of the "bread of shame" (a life built entirely of it leaves you wishing you'd found more nutrient-dense purpose).
- Values- or faith-aligned giving is a softer on-ramp than values-aligned investing: start with the philanthropic or donor-advised portfolio, where no financial return is expected, rather than fighting the "faith premium" debate inside the investment book — and note that even passive DAF capital is being allocated like any portfolio, so it can be aligned too.
- Separation can be its own success: one panelist exited the family business to find alignment and would count it a win even with little wealth created; another's outside ventures eased the tension with his father only once they began to complement, rather than compete with, the core enterprise.
- The family foundation as the next generation's training ground (one runs 34 years), alongside a daily meditation practice and silent family retreats — framed as breaking out of the business "prison" to create room for the next generation's freedom.
- Practical frameworks recurred: Maimonides' seven levels of giving; Chronos versus Kairos (clock time versus meaningful moments); the "true / kind / necessary" filter; and a 1,300-year, 46th-generation Japanese inn as the case study for stewardship over growth. The caution for founders: leaving only a "sell and give" instruction hands the next generation the task of defining what the giving is for.
Beyond the Surname: Forging Your Identity in the Shadow of Wealth
Panel discussion · speakers attributed by role
Headline takeaway
A next-generation panel on forging individual identity inside a wealthy family converged on a single reframe: succession is no longer about anointing one heir but about a "constellation" of roles — operators, board members, passive owners, and foundation leaders — matched to each person's genuine aptitude. A recurring argument was that the family foundation deserves to be treated as a legitimate, high-impact destination for talent rather than a consolation prize.
Key points
- The panel treated individual identity as something to navigate intentionally, not assume. Panelists described early entrepreneurial experiences — one starting at age 15 — that built confidence because the family created a trusting environment, with siblings on very different paths staying connected, and success defined in many forms beyond business metrics.
- Parenting was framed as decisive: it takes self-awareness to help a child be both an individual and part of the collective. The recurring metaphor was competing for a seat in the boat — the individualistic phase, with its sibling and cousin rivalry — then, once aboard, finding the shared belief that makes the boat go faster.
- Difficult family conversations can be cleansing when handled well. One panelist recounted an Asian family with two branches and in-law tensions that nearly separated, where having the hard conversation without it exploding revealed shared values — and surfaced the point that a family's intangible assets (communication style, shared experiences) are routinely undervalued against its financial assets.
- Succession and education come down to balancing freedom with responsibility: independence for the next generation alongside teaching them to be competent shareholders and board members. Many family members — doctors and other professionals — hold ownership stakes without basic business literacy, which takes an intentional, consultative, individually tailored approach to avoid.
- The "constellation" model was the panel's organising idea: replace the single-successor approach with multiple roles — active board members, passive investors, foundation leaders — through inclusive, transparent planning that goes well beyond the default of "good education plus outside work experience." The frameworks exist through business schools and family-business programs, and are well developed in the US but arriving more slowly elsewhere.
- On structure versus flexibility, founders surfaced a self-awareness problem: the entrepreneurial intensity that builds a business does not translate to running a family. One put it as "my job is to impose my will on the world" — which simply does not work at home — and called the first roughly 20 family meetings the most challenging part of the adjustment.
- Professionalizing family operations changes the dynamics: members behave differently once they see the full picture, the goal is authentic participation rather than deference to the founder, and the balance is between too much structure (rigidity) and too little (chaos). One founder described how hard it was to insist "it's not about me."
- A closing tension: legal structures set up 20-plus years ago can become today's constraints. The north star was that each generation needs room to make changes so the family isn't "running a museum" — holding legacy in balance with innovation, with trusted third parties helping navigate father-child tensions and the middle ground between individual happiness and preserving the legacy.
- A secondary thread on online identity and authenticity: owning how you show up online, and an LBS-stage interview whose message was authenticity — "be cringe, don't cringe" — set against the generational friction where next-gen members want an online presence their parents are deeply uncomfortable with.
Notable claims, calls, or numbers
- The "constellation" succession model (multiple roles matched to aptitude), set against the single-successor model, was the panel's central organising framework.
- A founder's candid admission that the first ~20 family meetings were the hardest part of professionalizing the family, because operating-business leadership style actively backfires in a family setting.
Disagreements or tensions
The disagreement here was internal and generational rather than between named experts: the pull between the next generation's desire for independence and the founder generation's instinct to protect through structure. Panelists also differed on how much structure is healthy — some emphasising clear development plans and opt-in expectations for board or ownership roles, others warning that over-planning and rigid legal protection can trap future generations and strip the flexibility that is itself a key family asset.
Implications for portfolio positioning
Not an investment session — there are no market or portfolio implications. The transferable governance insight for a family office is structural: design succession as a constellation of roles rather than a single line, build the literacy and opt-in pathways for shareholders and board members deliberately, keep legal and governance structures flexible enough that each generation can adapt them, and treat the family foundation as a legitimate, high-impact destination for talent rather than a consolation prize.
Speaker names were not captured for this session; contributions are attributed by role.
- The panel's reframe: succession is no longer about anointing one heir but a "constellation" of roles — operators, board members, passive owners, and foundation leaders — matched to each person's genuine aptitude, with the family foundation treated as a legitimate, high-impact destination rather than a consolation prize.
- Parenting is decisive: help each member be both an individual and part of the collective. The recurring image was competing for a seat in the boat, then, once aboard, finding the shared belief that makes the boat go faster.
- Difficult family conversations are cleansing when handled well, and a family's intangible assets — its communication style and shared experiences — are routinely undervalued against its financial assets.
- Build shareholder and board literacy deliberately: many family members hold ownership stakes without basic business literacy, and avoiding that requires inclusive, transparent, individually tailored planning rather than the default of "good education plus outside work experience."
- Keep governance flexible, because legal structures set up 20-plus years ago become today's constraints — each generation needs the ability to adapt them rather than "run a museum." Founders flagged a self-awareness trap: the entrepreneurial drive to "impose my will on the world" backfires at home, and professionalizing the family (the first ~20 meetings are the hardest) means trading deference for authentic participation. Not an investment session — the value is governance.