15 years on, I’m still asking why quantitative ERM struggles to catch on in emerging markets like Malaysia — when capable executives still choose coloured heat maps over a 370-year-old science of uncertainty.

The irony? The base tool for the upcoming IFRS S2 climate reporting is exactly that — ERM modelling.

Now companies must quantify (or explain why they can’t) how climate-related risks and opportunities affect future financial prospects.

Can ERM — in its current state — really handle this, or will it render IFRS S2 meaningless?

ERM must evolve, especially now that AI makes modelling far more accessible. But something keeps holding it back.

Over the weekend, I listed every systemic issue I’ve seen in ERM over 30 years — including the first 15 when I was a champion of heat maps — and ran them through a tool few in risk know: TRIZ.

What is this strange word?

TRIZ (Russian Teoriya Resheniya Izobretatelskikh Zadatch) means the Theory of Inventive Problem Solving.

I first came across it in 2010 while completing a quantitative ERM system we built from scratch — over 2 million lines of code — and asked:

“If risk assessment became rigorous and quantitative, would we just end up identifying problems but never solving them? How do we match risk treatment with the same intensity?”

Corporate problem-solving is still heuristic — driven by judgment, workshops, and PowerPoint. I wanted a system that could manufacture solutions, not just discuss them.

That search led me to TRIZ — used by Samsung, Daimler, Rolls-Royce, Siemens, and NASA. Created by Genrich Altshuller, a Soviet naval patent officer from Baku, who started analysing 200 000 patents since the 1940s, distilled 40 000 inventive ones, and found a universal pattern in the 1970s:

When one feature improves, another worsens.

He mapped these tensions into 39 system parameters — accuracy, adaptability, ease of use, speed, complexity. Think of it this way:

Make a car faster (Param #9 Speed) and you compromise safety (Param #31 Harmful factors).

TRIZ then points to a universal solution matrix — Taking Out (#2), Skipping (#21), Intermediary (#24), Parameter Changes (#35) — which led to innovations like lighter structures, airbags, and crumple zones — the logic behind safer, faster cars.

That same logic applies to ERM. When we make models more accurate, they become harder to use. And when we open them for standardisation and transparency, quants and consultants start worrying about IP protection.

So last weekend, I took the qualitative-versus-quantitative ERM debate for a TRIZ ride — to test whether a method born in engineering could solve the contradictions holding our field back.

I’ll share the results in pieces. Here's the first phase: mapping 44 ERM domain problems to TRIZ parameters, assisted by AI, test-checked against my TRIZ manual.