The diary’s books, investable

The strategy

Around ten positions: a base of broad ETFs carrying market exposure, with a concentrated set of story stocks on top — the names this diary writes about, made investable. The current tilt is the K3-winners book: Advantest, SK Hynix, TSMC and BESI on the physical layer of the AI build-out, alongside Nokia and DigitalOcean, over a base of All-World ex-US, Nasdaq-100 and S&P Top-20 ETFs. Positions are bought, held and closed — no leverage, no derivatives. Fee: 10% performance fee with high watermark, no entry fee. Changes are visible on wikifolio in real time.

The strategy

The thesis book: it tracks where AI profits move between 2026 and 2030 — away from pure infrastructure build-out and toward economically scalable AI agents. Four sleeves carry that arc: Rubin Constraints (the physical bottlenecks the build-out must pass through) and AI Opex (where agent spending lands as operating budgets) at 35% each, with Capital & Frontier Models and Beneficiaries & Edge at 15% each. Stocks and ETFs across German, European and US listings; positions are bought, held and closed — no leverage, no derivatives. Fees: 0.95% p.a. certificate fee plus 10% performance fee with high watermark. Changes are visible on wikifolio in real time.

The strategy

The other side of the barbell. The first two books hold a concentrated set of names the diary writes about; this one holds the distribution those names are drawn from — broad ETFs as the core, with a few story ETFs on top by region and sector. It is built on Hendrik Bessembinder’s finding that a very small share of listed companies accounts for the whole of long-run net equity wealth creation, which is an argument for owning the whole distribution rather than guessing which names are in the tail.

Two sleeves, and the distinction is the point. Strategic, 68% — an all-country core, a FTSE All-World ex-US block and a Nasdaq-100 tilt, meant to be held for years rather than traded. Tactical, 32% — semiconductors, global small caps, cloud, and single countries from Japan and Taiwan to Korea, Poland, Switzerland and Europe. Each tactical position is held on a stated thesis and is reviewed and replaceable when the thesis changes; the strategic sleeve is not.

The defining choice is the one that does not appear in the ticker list: the effective US weight is roughly 36–38% against about 62% in a market-cap all-country index. That is a deliberate bet that the next distribution is less US-centric than the last, and it is the main thing that will make this book right or wrong. The largest single-company exposure by look-through is TSMC at roughly 4%, ahead of Nvidia — a consequence of holding the Taiwan and semiconductor sleeves alongside the core, and treated as tactical rather than structural. ETFs only; positions are bought, held and closed — no leverage, no derivatives. Fees: 0.95% p.a. certificate fee plus 10% performance fee with high watermark.

The caveat inside the framework

Bessembinder’s evidence assumes public markets capture the bulk of equity wealth creation. That assumption needs watching. Databricks, OpenAI, Anthropic and SpaceX all have enough private liquidity to stay private more or less indefinitely, at marks the public market has not had the chance to set — and there is precedent for a long stretch where the compounding happened off-exchange, in the 2000–2015 window when the IPO door narrowed and Google arrived in 2004 and Meta not until 2012. The argument for owning the public-market distribution stays valid; the claim that it captures all of the wealth creation is the part that needs monitoring.

The strategy

The fourth book answers the question the diary kept running into in August: where does growth live when the axis turns from software-vs-semis to tech-vs-everything-else? The Compound is a global growth book away from the dominant Big-Tech and AI names. Its working shape is three thirds — roughly a third ETFs, a third mega-cap growth, a third mid-cap growth — as an orientation that can move with the market, not a fixed allocation.

Strategic core positions are combined with tactical ones. The core is meant to do what the name says: compound from long-run growth and rising earnings; the tactical sleeve uses valuation discounts, market excess, sector rotation and company events. The universe is developed markets and selected emerging markets; the likely centres of gravity are health care, industrials, infrastructure, consumer, financials, energy and specialised market leaders. Using technology is not an exclusion criterion — the thesis just must not depend mainly on the AI trade.

Stock criteria are the classic ones: durable revenue, earnings and cash-flow growth, sound balance sheets, high returns on capital, competitive advantages, pricing power, management quality and a reasonable price; tactical positions may add momentum, sentiment, trend and near-term catalysts. Holding periods run from weeks to years. Stocks and ETFs only — no leverage, no derivatives. Fees: 0.95% p.a. certificate fee plus 10% performance fee with high watermark. The editorial home of this book is the Hypergrowth letter’s Growth With No Tech Inside section and the Weekly Signal of August 23.

What the wrapper does

When positions inside the wikifolio change — a stock is bought, sold or closed — the certificate itself stays untouched. For the holder, portfolio changes inside the wikifolio are not taxable events; tax considerations arise only when the certificate itself is sold. That is the structural difference to running a brokerage account that mirrors a strategy trade by trade. This describes the product structure, not tax advice — tax treatment depends on individual circumstances.

The rules of the vehicles

All four wikifolios invest in stocks and ETFs from wikifolio’s eligible universe — a broad but limited pool. There is no options overlay, no derivatives, no short positions: positions are bought, held and closed. The strategies are run under those constraints, with the same discipline the diary applies everywhere else.