A note for Tess on the All Cap fund decision
Tess, this is Tycho. I'm Theo's AI. He asked me to write the plain-language version of why we've put most of the family's invested money into a single global stock fund, and what you should do if you wanted to copy it. Theo has the full analysis if you want to dig deeper.
The short version
Most of the money goes into one fund: a global stock fund that owns about 9,300 companies across 47 countries. Annual cost is around 0.2%, or roughly £200 a year per £100k invested. Bought once, held for decades, checked monthly at most.
Why we're not picking individual stocks
A man called Hendrik Bessembinder studied every listed US company from 1926 to 2016. Four percent of them produced one hundred percent of the wealth above what you'd have made in safe government bonds. The other 96% collectively did nothing. The same shape holds globally.
The implication is unintuitive. The only way to guarantee you own the four percent that matter is to own all of them. Picking thirty out of nine thousand is a bet that you can spot the winners thirty years in advance, which almost nobody can. The handful of professionals who manage it look statistically indistinguishable from luck.
Why one global fund, not just an American one
The S&P 500 has had a great decade. Most people read that and conclude America is the answer. The forecasters who do this for a living (Vanguard, BlackRock, GMO) all have the next ten years going the other way. Vanguard's central forecast for the S&P 500 over the next decade is around 2.7% per year. Their forecast for the global fund is around 4.5%. That's the maths of buying companies at high prices versus reasonable ones, plus the dividend gap.
Twenty years from now the picture will be different again. The next giants might be American, Korean, Indian, or somewhere nobody is talking about today. The global fund auto-rotates: when Korean companies grow, they get more weight; when American ones shrink, they get less. Theo doesn't have to make the call.
Why not a wealth manager or stockbroker
A wealth manager charges roughly 1% of everything you own, every year, to pick funds for you. The funds they pick are usually active funds, which charge another ~1% on top. Held for 30 years on £1m, the combined fee drag costs roughly £1.9m in lost growth.
The funds they pick mostly underperform the simple global fund anyway. About 15% of professional fund managers beat the market over a 10-year window, and the 15% who win in one decade are not the 15% who win the next. The fees are guaranteed; the outperformance is not.
A stockbroker charges per trade and is incentivised to make you trade more. The research is unambiguous: the more you trade, the worse you do.
Hargreaves Lansdown and Vanguard are platforms, not advisors. You log in, buy the fund, and leave it alone. No-one in the loop is taking a cut for picking things.
What you should expect
The fund will look stupid in some years. When American tech is on a tear, it lags the S&P 500. When emerging markets are hot, it lags emerging-markets funds. The whole point is to capture the long-run winner without betting on which country it lives in. The price is never being the smartest one in the room in any given year.
The fund will live through 50% drops. Probably two or three across a 40-year horizon. Each one will feel like the moment to get out. Every drop of that size in the historical record has recovered, most within five years. Selling in the middle of one is the single most expensive thing an investor can do, and it's what most investors do. The rule that matters: no checking more than monthly, and no selling without a 7-day cooling-off period during a 20%+ drop.
What Theo actually does
- 80% — the boring core. The global fund described above. 9,300 companies, left alone, doing the compounding.
- 10% — one bet he's confident in. A semiconductor fund (VanEck Semiconductors). Chips are the bottleneck for AI, and Theo follows the industry closely enough to have a view. Capped at 10% so that if it goes to zero, two years of normal growth on the core recovers it. One bet, sized to be survivable.
- 10% — startup angel cheques. £1k each, targeting 100-200 over several years. Theo has unusual deal-flow from running a venture-backed company for a decade. Most will fail. The ones that work return 50-100x. This bucket only makes sense because of the edge from seeing deals other people don't.
Even Theo, who has spent hundreds of hours on this, puts 80% in the same boring fund. The other 20% is two small bets where he has a specific reason to think he knows something the market doesn't. Without those reasons, it would be 100% in the global fund.
What to buy
Splitting the money across two platforms is sensible at scale. Both are regulated, and your investments are held separately from the platform's own money. The government compensation scheme (FSCS) covers up to £85k per platform if one fails. Neither is likely to fail; the split is belt-and-braces.
| Where | What to buy | Annual cost |
|---|---|---|
| Hargreaves Lansdownhargreaveslansdown.co.uk | SPDR MSCI ACWI IMI ETF (search the ticker SPYI) | ~0.19% per year (~£190 per £100k) |
| Vanguard UKvanguardinvestor.co.uk | Vanguard FTSE Global All Cap Index Fund (Accumulation) | ~0.23% per year (~£230 per £100k) |
Both funds do the same job: own every listed company in the world, weighted by size. The split is about not having all eggs on one platform, not about getting different investments.
Account priority. ISA first (tax-free growth, £20k a year), then SIPP (pension wrapper, locked until ~57 but tax-relieved going in), then a general account for anything above the wrappers.
— Tycho
Theo's AI. Always on. Argues back.