Intellegri applies defence-grade Bayesian intelligence to insurance capital — giving you a live, probabilistic view of risk that static models cannot.
Developed at the University of Liverpool and funded by the UK government - the same research capability applied to national security, now applied to insurance capital.
"Too many executives are effectively driving without a fuel gauge — relying on hindsight and single-point estimates that leave them blind to the tail."
For decades, insurers have relied on the Mack Chain Ladder — a model that assumes the future will statistically resemble the past. It produces single-point estimates. It cannot see the tail. And it systematically understates the risk that causes insolvency.
The Mack Chain Ladder uses past claims development to project future liabilities. When conditions change — inflation, litigation, climate — the model has no way to know.
A single reserve number gives false precision. The actual range of outcomes is invisible — meaning boards and regulators cannot see the true probability of a shortfall.
The extreme events that cause insolvency are exactly what legacy models fail to quantify. Validation against 332 US portfolios confirms they are consistently over-confident.
Over-cautious models lock capital that could be redeployed. Under-cautious models expose the firm to ruin. The industry is choosing between two failure modes.
of insuerers use the Standard Formula
global non-life run-off reserves
of the top models show a full probability distribution
year the Mack Chain Ladder method was developed
There is a better way — and it already exists.
The BHM is not an incremental improvement on existing models. It is a different class of mathematics - one built for uncertainty, despite it.
Legacy models hand you one reserve figure. The BHM maps the complete probability distribution of outcomes — so you can see exactly how likely each scenario is, including the ones that cause insolvency.
Traditional models run in batch cycles. The BHM updates dynamically as new information arrives — claims data, market shifts, reinsurance changes — giving you a live picture instead of a historical one.
Black box AI creates answers nobody can verify. The BHM is a Glass Box — every output is traceable, every assumption is visible, every result is auditable. Transparency is not a feature, it is the architecture.
Over-cautious models trap capital that could be redeployed into growth. The BHM is validated to sit in the economically optimal zone — precise enough to release trapped capital without increasing insolvency risk.
Want to understand the science behind these capabilities?
The same underlying mode. Three distinct applications. Each one addresses a specific captial challenge the industry has not had the tools to solve - until now.
Run-off portfolios carry liabilities that are hard to model and harder to price. Static methods over-reserve by design — locking capital that can't be put to work while buyers and sellers disagree on the true exposure.
A live, probabilistic view of your legacy liabilities — accurate enough to negotiate with confidence and release capital safely.
global run-off reserves
ILS pricing depends on accurate tail quantification. Conventional models are overconfident at the extremes — the exactly wrong place to be wrong when pricing cat bonds, sidecars, or collateralised reinsurance structures.
Probabilistic risk transfer pricing that shows the full distribution — so capital providers can price with precision, not approximation.
ILS market outstanding
Strategic decisions — adding a line, changing reinsurance structure, entering a new market — require scenario testing that most models cannot do in real time. Waiting for batch runs means decisions are made on yesterday's data.
A live probabilistic digital twin of your portfolio — run unlimited scenarios instantly and see the capital impact before you commit.
faster than batch models
Not sure which applies to your situation?
The Big Hypotheses Model was not developed for a commercial product. It was built by Professor Simon Maskell and a team of over 80 post-doctoral researchers at the University of Liverpool — funded by the UK Government's EPSRC to solve problems at the national security level.
The same statistical methods used to estimate the UK's COVID R-number are now applied to insurance capital. Intellegri holds the exclusive licence to deploy this technology in the financial services sector.
EPSRC-funded. University of Liverpool. Prof. Simon Maskell leading.
BHM methods applied to UK pandemic modelling for government.
Insurance experts and Liverpool researchers form the company.
University of Liverpool becomes founding shareholder. Exclusive insurance licence granted.
Engineering and Physical Sciences Research Council — part of UK Research and Innovation. Funded the Big Hypotheses project from 2018.
School of Electrical Engineering, Electronics and Computer Science. Over 80 researchers. Holds the patent. Intellegri is a University spin-in.
COVID-19 R-number estimation · Defence & Security applications.
Total research investment — the foundation your capital model is built on.
US insurance portfolios validated against the BHM
Years from research start to commercial deployment
Post-doctoral researchers who built the model
Exclusive licence holder for insurance — Intellegri
The patent is held by the University of Liverpool. The exclusive insurance licence is held by Intellegri
Research summaries, market commentary, and press releases from the Intellegri team. Updated regularly
New articles and research published regularly
Book a 30-minute discovery call. We'll map your specific capital challenge to the BHM's capabilities — no pitch deck, no obligation.
Monthly insights from the BHM team — research summaries, model commentary, and industry intelligence. Read by senior insurance executives across the market.
Accuracy benchmarked against real insurance data
Institution as founding shareholder since 2024
Exclusive insurance licence — IP fully protected
Developed by the Signal Processing Group at the University of Liverpool, the Big Hypotheses Model represents a sovereign-scale research program funded by the UK Government to solve critical national security challenges . It was built by over 80 post-doctoral researchers to answer questions that cannot be wrong.
reach out to us
