Helios Investment Journal · Taipei
Reading climate risk for investment portfolios, from Helios Investment Journal.
A free editorial journal on how machine-learning models are used to estimate climate exposure inside investment portfolios — written for readers in Taiwan who hold global and domestic assets and want to understand the method, not the hype.
What we cover
Climate risk for portfolios, examined as an editorial subject.
Most coverage of artificial intelligence in investing sells a verdict. This journal does the opposite: it slows down and reads the method. We look at how climate risk is translated into numbers a portfolio can carry — how a model turns a warming scenario into a line item that sits next to a company's earnings.
That translation has real limits. Emissions data is patchy, scope-3 boundaries shift between reporters, and forward climate scenarios are assumptions, not forecasts. The articles here trace those limits honestly so readers can judge which claims about AI-driven climate analysis deserve weight and which are decoration.
Nothing on this site is sold. No subscriptions, paid reports, managed accounts, deposits, trade signals or personalised investment advice are offered here — only free articles and an inquiry address for readers who want to ask a question.
Editorial principle
The useful question is not whether a model is “AI”, but whether its climate signal survives being explained to a sceptical reader in Neihu.
— the standard we hold each article to
How we read it
From a warming scenario to a number a portfolio can hold.
Climate risk analysis for a portfolio usually moves through three steps, and each one is where AI is either genuinely useful or quietly overrated. First, physical and transition scenarios are mapped to industries. Second, those exposures are pushed down to issuers and holdings. Third, the result is stress-tested against a portfolio's weights.
Machine learning earns its place at the seams — reconciling inconsistent disclosures, classifying business lines, and surfacing non-linear exposures that linear factor models miss. It does not remove the need for judgement; it changes where judgement has to be applied.
We write for readers in Taipei who manage global equity and fixed-income sleeves alongside domestic Taiwan holdings, and who need to see how a climate model treats a TSMC-weighted book differently from a European index fund.
Who this is for — and who it is not for
A journal, not a tip sheet.
This is for long-form readers: analysts, trustees, family-office staff and curious individual investors who want to understand how climate risk is modelled before they believe a number on a factsheet. If you want to follow the reasoning behind a climate-adjusted return, you are in the right place.
It is not for anyone seeking a trade, a rating, a recommendation to buy or sell, or a forecast of where a specific stock or index will be in twelve months. We do not produce those, and we do not sell access to anyone who does.
If a question falls outside an editorial journal's reach, we will say so plainly and point you to the kind of licensed professional who can help.
Inquiries
Have a question about an article, or a topic you want us to take seriously? Write to the journal.