Marc Kaio HaslerKaio Hasler
ZURICH SINGAPORE

I’m Kaio. I study finance in Zurich and write my thesis on LME industrial metals.

Marc Kaio Hasler, Swiss finance student. I go by Kaio. BA Banking and Finance at the University of Zurich, on exchange at NTU Singapore until December 2026. I build the tools I need with AI coding agents.

Available from January 2027 for a full-time internship in commodity trade finance, commodity trading or sustainable finance.

Email me LinkedIn Request my CV
ZURICH 47.4 N, 8.5 E SINGAPORE 1.3 N, 103.8 E

Why commodities

Why this is hereWhere my interest comes from, and why it is more than a trend choice.

It started with one course at UZH: Commodities Trading and Finance.

I have always liked finance best where it touches the real world: goods that are sourced, made, moved and paid for 1.

That is why Lavour, my school mini enterprise, mattered to me. We ran everything ourselves: sourcing, production, delivery and the financial planning behind it. I designed the wooden soap holder 2, and we milled it on a CNC machine my father and I built from scratch 4. A regional newspaper wrote about it 3.

You could say the first commodity I traded was soap.

Commodity markets and sustainable finance interest me, and I see the two as connected.

FIG. 1Trade in things that exist: metal that moves between warehouses.
The wooden soap holder designed for Lavour
FIG. 2The wooden soap holder I designed for Lavour.
A press piece about Lavour
FIG. 3Lavour in the press.
The CNC machine milling a wooden soap holderThe real machine, 2021
FIG. 4The CNC machine my father and I built to produce it.

My exchange: Zurich to Singapore

Why this is hereI chose both places on purpose. This is what I study there.

Switzerland and Singapore are two of the main hubs of commodity trading. I study in one and came to the other on purpose.

ZURICH SINGAPORE
ZRH, 2024 to 2027

Zurich, Switzerland.

BA Banking and Finance at the University of Zurich, 2024 to 2027.

  • Minor in General Economics
  • Bachelor thesis on LME industrial metals
Explore courses30 courses
Semester 1, autumn 2024
  • Informatics and the Economy3 ECTS
  • Business Administration I3 ECTS
  • Financial Accounting6 ECTS
  • Microeconomics I9 ECTS
  • Mathematics I6 ECTS
  • Banking and Finance I3 ECTS
Semester 2, spring 2025
  • Banking and Finance II3 ECTS
  • Business Administration II6 ECTS
  • Macroeconomics I9 ECTS
  • Mathematics II6 ECTS
  • Statistics6 ECTS
Semester 3, autumn 2025
  • Business Administration III6 ECTS
  • Introductory Econometrics6 ECTS
  • Microeconomics II3 ECTS
  • Introduction to Scientific Working3 ECTS
  • Corporate Finance6 ECTS
  • Banking6 ECTS
  • Startup Seminar Banking and Finance 2.03 ECTS
  • AI in Services Industries (seminar)6 ECTS
  • Summer school at WU Vienna: Social Media Marketing Strategy3 ECTS
  • Summer school at WU Vienna: International Business Strategy3 ECTS
Semester 4, spring 2026
  • Asset Pricing6 ECTS
  • Financial Economics6 ECTS
  • Financial Reporting3 ECTS
  • Macroeconomics II3 ECTS
  • Managerial Accounting3 ECTS
  • Commodities Trading and Finance (seminar)3 ECTS
  • Seminar Governance3 ECTS
  • Introduction to Health Economics3 ECTS
  • Corporate Entrepreneurship (seminar)3 ECTS

Request full transcript

Request full transcript

SIN, Aug to Dec 2026

Singapore.

Exchange semester at NTU, August to December 2026. Courses span commodity risk, fixed income, sustainable finance and working across cultures.

Explore courses4 courses
Exchange semester, autumn 2026
  • Enterprise Risk Management for Commodity Traders
  • Fixed Income Securities
  • Sustainable Finance
  • Cultural Intelligence at Work

An example of a tool I built for myself

Winning at Star Wars

Why this is hereA small example of how I solve my own problems: I build a tool.

Course registration at NTU runs on a system called STARS. Students call it ‘Star Wars’, because getting courses is a fight.

I built a course planner that optimises my course choices and gives me options and alternatives if I do not get a course.

Open the demo Demo data, not my real timetable

The path so far

Why this is hereThe short version of my CV, for context.

  1. Alte Kantonsschule AarauSwiss Matura, focus Economics and LawStudy2018 to 2022

  2. UBS ‘Most Innovative Product’ award 2021, top 8 of more than 180 teams nationwide.

  3. Led a group of about eight soldiers and designed a new reporting scheme. When the same exercise was repeated with it, reports were more precise and took about 60 percent less time.

  4. VistaioCo-founder and Head of OperationsWork2023 to 2025

  5. AI-assisted workflow automation in a B2B industrial company.

  6. Talks at Forum kAIpital and the International Bankers Forum.

  7. Reviewed alerts from the fraud detection model and worked with compliance and risk teams on flagged cases.

  8. Grade 5.3 of 6.0. Bachelor thesis proposal prepared: “Cancelled Warrants as Predictive Signals for Price Moves, Spread Dynamics, and Volatility in LME Industrial Metals”.

  9. Grade 6.0 of 6.0.

  10. Investment Club ZurichMacro Analyst for US marketsWorkSep 2025 to Sep 2026

  11. Courses include Enterprise Risk Management for Commodity Traders, Fixed Income Securities, Sustainable Finance and Cultural Intelligence at Work.

  • Germannative
  • EnglishC2 (TOEFL 114 of 120)
  • Frenchlimited working
  • Portugueseelementary

How I work with AI

Why this is hereAI is how I get work done every day. If you work with me, you should know how I use it and how I make sure the results can be trusted.

I keep all my work in one place, so the AI agent has the full context. People communicate better with context, and so do AI agents.

The principle

Sources first. Nothing imagined.

In everything I do, every statement and every number goes back to a source. Nothing is based on what a model imagines.

STATEMENT SOURCE

  • Sources are cited in APA style.

  • Every source is downloaded and kept.

  • Every number is traceable to its source.

  • A checker that uses plain logic, not an AI model, catches any figure that does not match its source.

The aim is to keep hallucinations and guesswork out.

From what I say to what gets built

Diverse minds avoid shared blind spots, so I never rely on one model.

  1. STEP 1
    I dictate

    What I want, in my own words. I dictate with a speech model from Nvidia that runs locally on my laptop and also works offline.

    WhySpeaking is more natural than typing, it leaves room for more context, and it is faster.
  2. STEP 2
    Three models sharpen the brief

    Claude turns the dictation into a clear brief. Gemini and ChatGPT interpret it independently and sharpen the task.

    ClaudeGeminiChatGPT
    WhyEach model reads a brief differently and has different blind spots. What one misses, another catches.
  3. STEP 3
    Claude plans and builds

    Claude Fable leads the planning, together with GPT and Gemini. Claude Opus does the work.

    Claude FableGPTGeminiClaude Opus
  4. STEP 4
    Three reviews, one logic check

    All three vendors’ models review the result independently. A logic check tests every number: plain logic, not an AI model.

    ClaudeGeminiChatGPTLogic check
  • As a second brain
  • For second opinions
  • For remembering small things
  • For building tools that help me every day
I built ymaplify with exactly this setup.

What I built

Why this is hereI learn by building. These are things I made because I needed them.

Live in production

ymaplify

Visit www.ymaplify.com
The need

When I travel I spend a lot of time on social media trying to find good places, and it is hard.

What it does

Paste a link from TikTok, Instagram or Google Maps and the place lands as a pin on your map. Trips are planned together on the same map. I built it for myself and my friends first. Now more people use it.

Why it can scale

It works like a small social network: anyone can make a map of good cafes or restaurants and share it with friends. It is also a personal record of where something was good and what you ordered.

ymaplify on a laptop: City map "My map of Singapore" from my own account: 75 saved spots, food filter on, pins across the city.
ymaplify on a phone: Same screen as laptop-1 in the mobile layout: map strip with pins, bottom sheet with the picks

01 / 03City map "My map of Singapore" from my own account: 75 saved spots, food filter on, pins across the city.

3D model
A property as a 3D model, rendered by Vistaio

01 / 04A property as a 3D model, rendered by Vistaio.

Visit the live website

Startup, 2023 to 2025.

Vistaio.

A startup I co-founded: 3D real-estate walkthroughs using Gaussian splatting.

“A 3D property modelling service with an interactive, game-like navigation experience.”

How Vistaio described itself on its website

The idea: scan a property with a consumer drone or a 360 degree camera, render it as a 3D model, and let buyers walk through it on a phone, on a desktop or in VR before they visit.

My part: operations, financial planning and investor materials.

We wound it down after market validation showed no clear product-market fit.

Bachelor thesis, proposal prepared

What I am researching, and why

Why this is hereWhat I am curious about, and how I approach a question.

Cancelled Warrants as Predictive Signals for Price Moves, Spread Dynamics, and Volatility in LME Industrial Metals

I find cancelled warrants genuinely interesting. My proposal is prepared. The formal thesis period is planned to start in late December 2026 or early January 2027. I wanted a topic that adds to the literature, that can be of interest to firms, especially trading firms, and that teaches me to work properly with statistics and with AI tools in research.

To my knowledge there is very little published research that tests this.

Beyond the thesis I also follow gold, silver and oil at the moment.

What I plan to test, in plain words

  1. 01
    What goes in

    Daily warehouse figures from the London Metal Exchange, cash and three-month prices for four metals, and a few market indicators such as interest rates and the US dollar.

  2. 02
    The signal

    Cancelled warrants: metal in LME warehouses that its owner has earmarked for withdrawal. I look at how much that is as a share of the stock, and how it changes from day to day.

  3. 03
    What it might tell us

    Whether a change in cancelled warrants says something about the next days: the price move, the gap between the cash price and the three-month price, and how strongly prices swing.

  4. 04
    How I keep it honest

    Every series is checked for gaps and coverage before any test. The LME data is licensed, so this page shows the method only: no data and no results.

Selected work

Why this is hereSo you can judge my work yourself instead of taking my word for it.

A selection of my papers and slide decks. Click a stack to turn its pages, or open it large to read.

Seminar Commodities Trading and Finance, University of Zurich, 2026

Group of five
Page 1 of 33, The Enron Collapse: mark-to-model trading, synthetic leverage and the limits of disclosureTurn page
Page 1 of 33
01 Paper, 33 pagesGroup of five

The Enron Collapse: mark-to-model trading, synthetic leverage and the limits of disclosure

Seminar Commodities Trading and Finance, University of Zurich, 2026

A 33 page case study written by a group of five.

Slide 1 of 10, The Enron Collapse: pitch deckTurn page
Slide 1 of 10
02 Slide deck, 10 slidesGroup of five

The Enron Collapse: pitch deck

Seminar Commodities Trading and Finance, University of Zurich, 2026

The presentation that goes with the case study.

Seminar Governance, University of Zurich, 2026

With one co-author
Page 1 of 4, Open letter to the management of Glencore PlcTurn page
Page 1 of 4
03 Paper, 4 pagesWith one co-author

Open letter to the management of Glencore Plc

Seminar Governance, University of Zurich, 2026

A four page letter on Glencore and ESG activism. In German.

Slide 1 of 10, Glencore and ESG activism: discussant slidesTurn page
Slide 1 of 10
04 Slide deck, 10 slidesWith one co-author

Glencore and ESG activism: discussant slides

Seminar Governance, University of Zurich, 2026

Ten slides for the discussant talk on Glencore. In German.