About Adrian Clements

About · Essay · Risk & Decision

CORE Creation — From Info to Decisions

Every crisis taught the framework something. Thirty years of building risk engines — from a reinsurer's claims history to CORE©. A note on how the ideas grew into a workable product.

Author Adrian Clements
Published November 2026
Read ~8 minutes

Risk management has a quiet failing. It tells you, with real precision, what might happen — and almost nothing about when you must decide, what it will cost you to be wrong, or what it means for the company's future performance. I have spent thirty years narrowing that gap. Each tool I built solved the problem in front of it and left one idea behind. Set end to end, those ideas became CORE©. This is how they grew.

Thirty years, six engines, one method

01
The reinsurer's engine HELGA

Aggregation, and the tail

My first engine simulated a client's losses across every line of business at once, then sized how much risk they could safely keep and the capital needed to stand behind it. The lesson arrived immediately. Capital is not driven by the average claim; it is driven by the aggregate and its tail.

On one chemicals book, a single plant fire was sixty percent of seventeen years of losses. On a banking group, the largest insured losses were the vehicle fleet, not the vaults. And pooling retained risk into one financial transfer vehicle, rather than funding each line alone, cut the capital required by close to half. Risk, it turned out, almost never sits where you expect it.

The idea it added

Model the whole, not the parts — and respect the tail, because that is what you are actually holding capital against. Be aware of small risks that can aggregate into a sandpile effect.

02
Into industry HERA

Exposure as one number — and time

Reinsurance looks backward, at losses that have already happened. Industry needed to look forward, at exposure that has not. So I built HERA, which compressed asset vulnerability — weighted by how often something fails and how quickly it bites — into a single index you could rank real capital expenditure against.

The idea it added

Make exposure a decision-grade number, and treat time-to-impact as part of the score, never a footnote.

03
The platform Christobal

Appetite you can see

A score on a spreadsheet changes nothing. HERA became Christobal, a platform running across more than 150 industrial sites, where risk appetite was drawn as a curve on the chart and every site could be compared on one scale. You navigated it by clicking the data, not by filtering menus.

The idea it added

A method creates value only when it becomes a place people work — and when the boundary between acceptable and unacceptable risk is visible on the page. Insight came from understanding how each action affected severity, frequency or both, and to what degree. Without this we don't know the risk direction.

04
The turn Stability Index

Volatility is the signal; EBITDA is the point

Then the step that changed everything. Using live production data, I built the Stability Index, which treated equipment reliability as a proxy for the quality of risk management itself. It read each site on three things — the spread of its industrial processes in terms of target achievement, their volatility, and their trend — and, borrowing a momentum measure from the financial markets, watched whether reliability was turning up or down before the raw numbers made it obvious.

The reliability spread then became a confidence band around the EBITDA forecast. Two ideas arrived together. The volatility of the asset base is itself the risk signal; and reducing it is precisely where cash is made. Risk management and financial performance were, at last, the same conversation.

The idea it added

Risk is performance. And to catch the turn early you need an unbounded measure — one that keeps rising as momentum builds, not a score quietly capped at ten.

05
Adding time HEROD

When, not just how bad

The last missing dimension was time itself. HEROD added velocity and acceleration, the way one risk cascades into others, and a plain decision grammar — probe, hedge, commit — set inside a diagram borrowed from physics that shows the exact moment a decision window closes.

The idea it added

The timing of the decision and its realisation matters as much as the severity of the risk — and risk and opportunity are the same mathematics, read from opposite sides. Slow decisions cost money, and being good at risk management does not mean you are good at opportunity management.

06
The synthesis CORE© · RADAR© · FORGE©

Everything, at once

CORE© is the first version that carries all of it — simulation, exposure, platform, volatility, timing — in a single method. CORE© holds the mathematics. RADAR© watches for what is emerging, unbounded and sensitive: the Stability Index's momentum, grown up. FORGE© governs the response, bounded and stable. Nothing from the thirty years was thrown away.

Each crisis had taught the framework something. The framework remembered.

Frameworks are usually handed down as finished things. Mine was assembled, slowly, out of real books of risk, real plants, and real decisions that could not wait for more information. That is the only reason I trust it.

— Adrian Clements

Where the thirty years led

Read the CORE©, RADAR© and FORGE© white papers — or the Insight Series that puts them to work.

Read the CORE© Paper Read the Blog