hcp.ch / the-idea
The idea behind HCP.
Not another product shelf — a machine-learning stock-selection engine, built in-house over several years, applied where it's actually been tested.
Please click 'Agree' to the following disclaimer to continue reading, or email us at contact@hcp.ch for any questions:
Declaration of consent
By using this website, you confirm that you are a professional/institutional client respectively a qualified investor domiciled/resident in Switzerland and that you agree to the following terms and conditions of use.
Access and restrictions
The website is intended solely for persons that are resident in or have their registered office in Switzerland. Persons subject to a jurisdiction that prohibits access to the website due to their nationality, place of residence or domicile, or for other reasons are not permitted to access the website. Furthermore, some services and products described on the website are subject to restrictions on the basis of place of residence, domicile, nationality or client segment. Therefore, certain services and products described on the website may not be permissible or suitable for you.
No advice and no offer
The information published on the website is for information and advertising purposes only. It does not constitute an advice or – from a legal, tax, financial or other point of view – a recommendation or offer to purchase or sell products or to conduct or conclude any other kind of transaction or legal act. The fact that you are accessing the website does not make you a client of HCP Asset Management SA.
Exclusion of liability
HCP Asset Management SA exercises customary due care in creating information published on the website. Nonetheless, HCP Asset Management SA does not accept any liability, in particular for the up-to-dateness, correctness, accuracy or completeness of all the information on the website. HCP Asset Management SA reserves the right to amend or delete the information on the website at any time without notice.
To the extent permitted by law, HCP Asset Management accepts no liability whatsoever for direct and/or indirect loss or damage or missed profits arising as a result of using or accessing the information on the website.
Reservation of all intellectual property rights
All information on the website is protected by intellectual property rights and belongs to HCP Asset Management SA or to third parties. The information may be freely used for private purposes only.
Right of amendment
The terms and conditions of use, the legal information and all information on the website may be amended at any time without prior notice. If you do not understand or do not agree with a provision in the terms and conditions of use, please leave the website.
Reproduced from the live hcp.ch disclaimer (last updated April 2024) — abridged in the liability/IP sections for this preview, condition-for-condition unchanged.
While the market talks about LLMs, HCP spent the last several years building something narrower: a machine-learning engine whose only job is picking stocks.
Built in-house, not bought
HCP's stock-selection engine is proprietary machine learning developed internally over multiple years of research — not a licensed model or a marketing wrapper on someone else's LLM.
A narrow job, done deliberately
Not a general-purpose AI assistant — a model trained specifically to rank and select equities, independent of the current LLM cycle.
Tested, then extended
Applied first to European large caps (the Euro Stoxx 50), then extended into new US-market products — a track record before a launch, not the reverse.
HCP doesn't publish the engine's methodology or current performance on the open web — per their own site, that detail is shared directly: contact@hcp.ch.
Positioning
Why HCP, relative to the alternatives
Draft — a real comparison needs HCP's actual peer setvs. a private bank
No in-house products to place — the incentive is the mandate's performance, not this quarter's structured-note shelf.
Trade-off to be honest about: less balance-sheet, fewer ancillary services (lending, custody-in-house).
vs. a robo-advisor
The stock-selection itself is systematic — the same ML engine end to end — but a licensed team still makes the call on a client's actual portfolio, not a questionnaire-assigned model.
Trade-off to be honest about: higher minimums, less automation, human-paced onboarding.
vs. a large asset manager
A stock-selection engine built and owned in-house, not licensed from a data vendor — and a mandate that's a relationship, not an account number.
Trade-off to be honest about: smaller research bench, narrower strategy shelf.