Portfolio Part 2: The core I meant to build
How one deliberate pick grew into a quarter of everything, and the four years I barely contributed
Where part one left off
Part one covered the money I put in during the pandemic: about CHF 2,800 into Rolls-Royce that is now worth roughly CHF 35,000, a banked profit on Siemens Energy, and five other positions that between them came to a small net loss. Two ideas carried the whole era.
This part is about what that left behind. The account is worth approximately CHF 143,000 against CHF 77,000 of cash deposited, and the interesting question is no longer how the winners did. It is what the portfolio actually looks like now, how far that is from the thing I meant to build, and why I spent four years barely funding it.
What the portfolio actually looks like
Here is the part that took me longest to see properly. This is what I paid for each current holding against what it is worth today.
Current holdings, rounded CHF. Cost is net of any capital returned.
Almost every line is short. Money went in, a bit more came out, which is what a functioning portfolio looks like over a few years. Then there is Rolls-Royce, stretching most of the way across the chart on its own: 3.1% of every franc I have ever invested and 24.5% of what the portfolio is worth today.
Rolls-Royce is a quarter of the portfolio from 3% of the money.
Nothing about the size it grew to was a decision. I chose to buy Rolls-Royce; I never chose to let it become a quarter of my portfolio - price growth did that while I was looking elsewhere. The risk does not care how deliberate the original pick was. If Rolls-Royce halved tomorrow, the account would lose more than the entire cost basis of my ten largest deliberate purchases combined.
Not everything recent has worked, either. Novo Nordisk is my clearest current loser - about CHF 13,700 in, worth roughly CHF 11,200 now, down some 18% since I started buying in April 2025. SAP is modestly underwater too. Being deliberate is not the same as being immediately right, and I would rather show those two lines than quietly crop them out.
The core I meant to build
The structure I actually want is not complicated. A broad, diversified ETF core should do the heavy lifting - global trackers and a few income funds, bought regularly and more or less ignored. Something like 60/40 in favour of the funds, and I would be happier at 70/30. The individual companies were supposed to be the satellite: a deliberately small slice of fun money, there to keep me interested enough to keep paying attention to the boring part. That is not a throwaway benefit either - money that is invested and being watched is money I am far less likely to quietly spend, and there is a genuine bit of entertainment in checking a share price or arguing myself into the next purchase, in a way there just is not in watching an index fund do the same 8% it always does.
That is the theory. In practice it looks rather different.
- ETF core38.3%
- Rolls-Royce24.5%
- Other individual companies37.2%
Share of portfolio value today.
The fun money is 62% of the portfolio, which makes it something closer to the main event than a satellite.
Some of that is one position running away with the scoreboard rather than a habit of picking. Set Rolls-Royce aside and, measured by money actually spent rather than value today, the split comes to 51% core, 49% satellites. But 50/50 was never the target either - I wanted 60/40 at worst. So even by the measure that strips out the twelvefold return and asks only what I chose to buy, I have been overweight individual companies for the entire life of this account, and that part is on me rather than on how one holding happened to perform.
The fix is not to stop picking companies. It is to get the ratio to 60/40 and keep it there, which is an arithmetic problem rather than a character flaw, and one that adding money solves without my having to sell anything.
I should admit the awkward part, though. The satellite was meant to be the risky bit I could afford to lose, and instead it has outperformed the core by a wide margin - on the strength of reasoning I have already conceded was not rigorous. Two good outcomes out of seven attempts, with one of them enormous, is exactly what survivorship bias looks like from the inside.
The gap in the middle
The deposit history tells a story the value chart hides.
2026 covers January to September. 2024 is not a rounding error.
For four years I added a few thousand francs annually, tapering to CHF 2,500 in 2023. Then, in 2024, I added nothing at all. Not one franc.
The plain explanation is mundane: 2024 was the year I cleared my student loan, alongside the usual run of expenses that arrive whether or not you have a plan for your brokerage account. Paying off debt instead of buying shares is a defensible choice, and I do not regret it.
What I do not get to claim is that the whole four-year drought was that deliberate. Long before 2024 I was already contributing thinly and irregularly, because I had no mechanism - investing was something I did when I happened to think about it and happened to have money spare, so it competed with every other call on my attention and usually lost. A year of large expenses on its own would not have produced a blank year - it took that plus having no standing order in place to make it one.
The result either way is that the account sat there for twelve months being carried entirely by a position I had bought and then left alone since 2022.
The plainest way to sum up my first four years is that I tried some things out. I set up what I meant to be a core, made a few bets alongside it, and then let it sit while I got on with other things. The intentions were sound and the structure was roughly the right idea; what was missing was any process for continuing to act on it.
What broke the drought was deciding to treat this as a plan rather than a side project. In 2025 I deposited CHF 35,000, and CHF 24,600 more in the first nine months of 2026 - more in twenty months than in the previous four and a half years combined.
That money went in deliberately and sometimes quickly. April 2025 alone accounts for about CHF 19,000 of purchases, made in the weeks after the tariff-driven sell-off: ASML and Novo Nordisk both started there. September 2025 and January 2026 saw two more large deployments into the Swiss dividend ETF, the global high-dividend fund, SAP and TSMC. Buying quality companies into a panic on purpose, rather than stumbling into one, is the single biggest change in how I invest.
Rounded CHF estimate from month-end holdings and prices.
The reconstruction is an estimate built from transactions and monthly market data, not a broker statement. But the shape is right: a flat, quiet middle, then both lines climbing steeply from 2025 as real money started arriving.
Finding a rhythm
The deposits above are lumpy for a reason, and it deserves unpacking because it is the part I am now trying to make deliberate. Plotting every deposit by the month it landed shows a pattern that was invisible in the annual figures.
Every deposit by calendar month, split by era.
Sixty-five per cent of everything I have ever deposited arrived in January, August or September. In the early years the deposits were scattered thinly across the calendar. In the current era they arrive in two concentrated bursts, and both have a cause.
January is the bonus, when there is one. It is tied to company and divisional performance rather than guaranteed, but when it lands it is usually the largest single inflow of the year.
August and September are stranger, and they are a Swiss tax artefact. On a C permit I am no longer taxed at source, so I owe a real tax bill and pay into a separate tax account monthly to meet it. Estimating that conservatively means I usually overshoot, and the rebate arrives in late summer. That money was never really spare - it is my own over-payment coming back - but it lands at a point in the year when I already know my tax position for certain, which makes it the safest surplus I get.
The monthly drip is the newest habit and the one I am worst at. A standing order does not care whether the market looks expensive, and more to the point it does not care whether I have decided what to buy - which is the entire point, because 2024 proves that is the decision I am worst at making. The two annual lumps are where the real money moves, but they are also where timing risk concentrates: deploying a bonus in a single January morning is a bet on that morning, and spreading it across a few weeks is something I should probably start doing and currently do not.
One more thing explains why the monthly number is small in the first place. As I covered in part one, the shares account is the last claim on the money rather than the first: roughly CHF 1,000 a month goes into Pillar 3 and a further voluntary slice into Pillar 2 before anything reaches the broker at all.
What the monthly number is actually worth
The reason I keep returning to the monthly contribution - rather than to stock picking, which is far more fun to write about - is that it is the variable with the most leverage over how this ends. The controls below are live, so you can put your own assumptions in rather than take mine.
- NothingCHF 2.06M
- CHF 1k/monthCHF 3.75M
- CHF 2k/monthCHF 5.43M
- CHF 3k/monthCHF 7.11M
- CHF 4k/monthCHF 8.79M
Compounded monthly, with contributions stopping at retirement. The dotted line marks that point; the 5 years beyond it are the same money left alone. Legend figures are the balance at retirement, the labels on the chart are where each line ends. Figures are in today's francs and assume a constant return, which no real portfolio delivers.
On the default settings - 8% growth plus a 2% dividend reinvested, twenty-eight years to run - today’s CHF 143,000 becomes about CHF 2.1 million on its own, without my ever adding another franc. Compounding does the overwhelming majority of the work.
But look at the gaps. Adding CHF 1,000 a month takes it to roughly CHF 3.8 million. The contributions themselves total CHF 336,000, yet the end result is about CHF 1.7 million higher than doing nothing - the deposits are worth five times what I actually pay in.
Turn the dividend reinvestment off and the same CHF 143,000 grows to CHF 1.2 million instead of CHF 2.1 million - a fairly blunt argument for the dividend-growth strategy in part three.
Run the same numbers at UBS’s published savings rate instead of the market - the same 0.05% used in the deposit chart above - and CHF 1,000 a month for twenty-eight years barely clears CHF 340,000, most of it the cash itself. That is the actual choice being illustrated here: not shares against a slightly worse version of shares, but shares against an account that cannot outrun inflation at any horizon.
The five years drawn past the dotted line make the same point from the other direction. Nothing is paid in over that stretch, and the do-nothing line still climbs from CHF 2.1 million to CHF 3.3 million. Those five untouched years add about CHF 1.2 million, more than three times the CHF 336,000 I would have deposited over the preceding twenty-eight years at CHF 1,000 a month - which is why the retirement age is an input rather than a fixed 65: moving that date is one of the larger levers available, and it costs nothing to try.
Those defaults are not plucked out of the air, but they are not neutral either. Vanguard’s FTSE All-World ETF, priced in francs on the Swiss exchange, has returned 7.9% a year in price and 2.1% a year in dividends since 2013 - almost exactly the 8% and 2% the chart opens on, and already net of the currency drag that eats a Swiss investor’s global returns. The problem is the window: thirteen years starting in 2013 contains no lost decade, and the long-run record for global equities is closer to 5% a year after inflation. At 6% growth plus the 2% dividend, CHF 1,000 a month lands at about CHF 2.4 million rather than 3.8, and doing nothing leaves CHF 1.2 million rather than 2.1. The return assumption moves the answer further than my own behaviour does, and it is the one variable I have no influence over whatsoever - which is the argument for dragging the growth slider down and looking at the pessimistic version.
I should say plainly what the model is not, too. A constant annual return is a fiction: the average conceals crashes, and the sequence of returns matters enormously if the bad years arrive late. It ignores inflation, so those figures are in today’s francs made to look like tomorrow’s. And it assumes I keep contributing through every downturn, which this post has already shown I am not reliably capable of.
It is still the clearest argument I have found for automating the contribution. The difference between the lines is not skill, insight or picking the right jet-engine manufacturer during a pandemic. It is just the standing order.
The constraints nobody writes about
Portfolio posts tend to describe an optimum. What they leave out is that the thing you end up holding is shaped as much by what was available and what was going on in your life as by what you decided.
The broker decides some of it for me. I hold TSMC through an American depositary receipt rather than the Taiwanese listing, because that is what DEGIRO offers - a small, slightly annoying compromise of exactly the kind that never survives into a tidy strategy write-up.
Life gets in the way. The plan is a steady monthly contribution. The reality is that 2024 went on clearing a student loan, and other years have had their own version of that. A standing order does not remove the competing expense, but it does stop the shares account being the thing that silently loses every time.
Political volatility cuts both ways. It has been good to me so far - the pandemic crash created the Rolls-Royce position, and the tariff-driven sell-off in April 2025 is when I bought ASML and Novo Nordisk. But the same instinct that says “buy the panic” is indistinguishable, in the moment, from the one that bought two airlines. A falling price is not a thesis, and the periods that have rewarded me are exactly the ones I should be most suspicious of.
Where that leaves the plan
Two problems come out of all this, and they are different in kind.
The first is structural: 62% of the portfolio sits in individual companies when I wanted 30-40%, and a quarter of everything rides on one British engine maker. That is a fixable arithmetic problem and the fix is unglamorous - keep adding to the underweight side until the percentages move.
The second is behavioural, and it is the one that actually decides how this ends. For four years I had no mechanism, so contributing competed with everything else and usually lost. The projection above puts a number on what that costs, and the number is large enough that no amount of good stock picking would have made it back.
What neither of those tells me is what to buy with the money once it is arriving reliably. The pandemic method - read the news, reason about a trend, hope the business survives - got me one excellent outcome and two write-offs, and I would rather not run it again. Part three is about the strategy I am borrowing instead, why I am bending it away from its US-centric original, and whether the numbers say it is working yet.