Swiss bank mergers are the founding myth of this industry, and everyone in it has a version of the same origin story. The bank merged. The role disappeared, or the culture did. The retention package ran its course. And two years after the deal was signed, a new independent wealth manager opened its doors.
It is a good story, and it fits what I have watched happen around me for twenty years. So I tried to find the Swiss bank mergers in the numbers.
It is there once, faintly. What is far clearer in the data is something the story does not mention at all.
What the chart cannot tell you about Swiss bank mergers
Let me deal with the problem first, because it is the whole reason this piece is not a triumphant confirmation.
Every firm in this dataset exists today and holds a licence. That is what a FINMA list is. Firms founded in 1999 that closed in 2013 are not here. Firms founded in 2019 that closed last year are not here either, but there has been far less time for that to happen.
The consequence is that the rising slope from 1995 to 2019 is partly real growth and partly the shadow of thirty years of attrition. I cannot separate the two from this source. So any statement of the form “the industry produced five times more firms in 2010 than in 1995” is not supported, and I am not going to make it.
What does survive the bias is a sharp break between adjacent years. Two neighbouring cohorts have had almost the same time to die, so a jump from one to the next is a real change in behaviour, not a survivorship artefact. That narrows what we can read to a handful of moments — but those moments turn out to be the interesting ones.
What the Swiss bank merger data can tell you
2018 is the tallest bar in the series. 45 firms in 2017, 78 in 2018, 58 in 2019, then 31 in 2020. That is not a bank event. FINIG was adopted in June 2018 and came into force on 1 January 2020, and firms already active before that date kept the right to operate while their application ran. 2018 and 2019 were the last two years in which you could start first and get authorised later. The spike is a queue.
2009 is the second tallest. 58 in 2008, 70 in 2009, 64 in 2010, 61 in 2011. UBS was rescued in October 2008; the US settlement and the end of banking secrecy as we knew it followed in early 2009. This is the closest the data comes to the Swiss bank mergers story, and the timing has exactly the lag you would expect — the effect arrives the year after the shock and stays elevated for three more. Twelve firms is a real jump in a series that otherwise moves by five or six, but I would call it suggestive rather than settled.
Note what that lag implies. People do not leave during the crisis. They leave once the outcome is known, the restructuring is announced, the role is defined or removed, and the deferred compensation has either vested or been written off. The independent firm is founded when the guarantee expires, not when the deal is signed.
One bar I cannot account for: 2016, with 61 firms, sitting between 42 in 2015 and 45 in 2017. No bank event, no regulatory deadline that I can tie to it. I flag it because a chart with three explained spikes and one unexplained one is a more honest chart than one with three.
The Credit Suisse non-event
Which brings us to the case everyone assumes is obvious, the largest of all Swiss bank mergers.
UBS agreed to take over Credit Suisse in March 2023 and completed in June. By the logic above, the founding wave should land in 2025 and 2026, as retention periods run out and integration decisions become final. It should be the largest such wave in Swiss history, because it was the largest such event.
In this dataset, 2023 and 2024 together produced 48 new firms. 2018 and 2019 together produced 136.
2025 and 2026 show zero, and I am going to set that aside rather than lean on it. A firm cannot appear on a list of licensed firms until it is licensed, and under the current regime that has to happen before it takes its first client. A zero in the two most recent years is what the definition produces. It is not a finding.
So the CS wave is not visible. That is worth saying plainly, and it is where most versions of this analysis would stop. But the chart contains something that the Credit Suisse question distracts from, and it is the more important of the two.
The number that actually matters
Look at the five years since FINIG came into force, well after the big Swiss bank mergers. 31 firms in 2020, 22 in 2021, 35 in 2022, 26 in 2023, 22 in 2024.
That is 136 firms in five years. 2018 and 2019 produced 136 firms in two.
The decade before FINIG ran at roughly 57 foundings a year. The five years after it run at 27. The rate did not dip and recover. It halved and stayed there, through a pandemic, a boom, a rate shock and the largest bank failure in Swiss history, without moving much either way.
This is also the one comparison in the chart that survivorship bias cannot take away from me, because here the bias runs the wrong way. Older cohorts have had thirty years to lose members; the 2020s cohorts have had barely any. If anything the recent years are flattered and the older ones understated. The real gap is at least this large, and probably larger.
There are two readings of what the Swiss bank mergers left behind, and I cannot yet choose between them.
The first is backlog. FINIG turned authorisation from a formality into a precondition, applications took years to clear, and the pipeline is still draining. On this reading the rate returns to something near its old level once the queue empties.
The second is that the wave is being redirected rather than delayed. Founding your own shop is no longer the default exit from a bank, because it now requires a licence, a supervisory organisation, an organisation and capital before revenue. The banker who in 2009 would have incorporated joins an existing firm in 2025, or a platform, instead. The wave still happens — it just does not show up as new companies.
Five years is a long time for a backlog. I lean towards the second reading, and I hold it loosely.
A prediction I am willing to be wrong about
These two readings of the Swiss bank mergers make different, checkable forecasts.
If it is backlog, the FINMA list published in 2028 will show 2023, 2024, 2025 and 2026 filling in towards the old 55-per-year level, and the cliff in this chart will flatten out.
If the route has genuinely narrowed, those years will stay near 25 to 30 even after the queue clears, and the CS departures will be visible only in headcount at existing firms — never in the founding statistics at all.
Same list, same method, two years from now. I will run it again and publish the answer either way.
And the stock market?
Briefly, because the honest answer is short. The founding series does not show the equity cycle cleanly. There is no visible dip in 2002 after the dot-com crash, and no visible dip in 2009 — if anything, the opposite. Whatever drives a Swiss banker to go independent, the level of the index in the month they decide is not it. Employment shocks — Swiss bank mergers above all — show up in this data. Market shocks do not.
Method. FINMA list of licensed portfolio managers retrieved 21 August 2026; commercial-register entry years established for 1,329 of 1,358 firms. The dataset contains only firms licensed and existing at the retrieval date, so it is a survivor sample: cross-decade level comparisons are unreliable, adjacent-year changes are not. Event dates are matters of public record. The register year approximates the founding of a wealth manager well after 2000 and poorly before it.
