The risk is already
on the balance sheet.
The question is whether it was decided or inherited.
Risk is inherent to any operation. Every company carries risk on its balance sheet. The decisive question is: how much of that risk was actually decided? How much remains after engineering has done what it could and transfer has covered what it should?
Retention Engineering is the discipline built for this intersection. It measures, models and governs the share of risk a company effectively retains, integrating the three readings into a single capital-decision agenda.
In practical terms, the method answers three questions: what is the plausible severe loss, how much of it remains on the balance sheet after effective recovery, and whether that exposure fits within the company's prudential capacity.
The CFO speaks the language of the balance sheet. What is missing is an indicator that names the risk the balance sheet already carries.
How much capital the balance sheet can commit to an adverse event before it triggers a cash constraint, a covenant breach, a loss of liquidity or the need for an injection. Not an insurance indicator. A financial indicator that is rarely made explicit in risk-transfer decisions.
How much of the retention sitting on your balance sheet today was actually decided?
In an initial conversation, Retention Engineering does not start with a model. It starts with a simple reading: which loss could truly affect cash, capital and continuity; how much of that loss would be effectively recoverable; and how much would remain on the balance sheet.
An initial conversation can run on public information alone. The objective is not to issue a diagnosis, but to test whether there is a relevant retained exposure to be governed.
Test the initial readingThe company already understands risk. It just reads it in three languages that do not speak to one another.
Risk transfer sees what leaves the balance sheet. Finance sees what the balance sheet can absorb. Engineering sees what can happen. Each reading is competent in its own field. What changes everything is when the three meet.
What leaves the balance sheet.
Policy, limit, contracted retention. Looks at cover and at premium. A competent reading of what has been ceded to the market.
What the balance sheet can absorb.
Capital, liquidity, covenants. The CFO reads exposure to volatility and to the cost of shock. A competent reading of absorption capacity.
What can happen.
Scenarios, severity, failure mode, recovery time. The plant reads the technical event and the cost of the loss.
When the three readings meet, retention stops being a consequence of the policy and becomes a capital decision.
At the centre, the share of risk the company effectively carries — after engineering, after transfer, within the absorption capacity of the balance sheet.
What remains with the company after the contractual response.
Cover on paper is not cash on the day of the loss. And not every loss is transferable. ERE reads what is left standing in the company once the policy has returned what it could, plus what was never covered to begin with.
Three readings call for three answers. The answers balance against each other.
The indemnity restores the asset in fourteen months. Who pays the payroll during that time?
Contracted cover is not the same as cash available on the day of the loss.
Between the event and the indemnity there is an interval. It can last months. It can last more than a year. During that interval the operation continues. The payrolls continue. The contracts continue. The covenants continue.
The asset loss is restored. The liquidity is not. That is why the retention decision has two axes, not one.
The boundary between the three answers is a capital decision.
Calibrating it is the work of Retention Engineering. Each vertex carries its own rationale. Every move on one vertex changes what remains for the other two.
Retain
Size calibrated against absorption capacity. Explicit band. Recorded criterion. Periodic review.
Transfer
Insurance programmes, captives, parametrics, guarantees, contracts. Layer and deductible adjusted to what is worth ceding.
Mitigate
Reliability engineering, redundancy, maintenance, operating controls, segregation, protection, detection and response. Reduces probability, severity or recovery time.
Absorption is not a single number. It is a decision scale.
The three answers live in three different planes. Each plane answers a different question about the same balance sheet.
The map.
Size calibrated against absorption capacity. Explicit band. Recorded criterion. Periodic review.
The route.
Physical mitigation, contractual transfer, contingent liquidity, guarantees, hedges, governance adjustments. Reorganises what fits in each lever.
The boundaries.
Financial clauses that do not reduce the event. They define when a loss ceases to be absorbable and becomes a capital crisis. Headroom, triggers, breach.
The three planes do not compete. Retention Engineering articulates the three into a single decision agenda.
The discipline was born from practice.
More than three decades working in high-risk industries made one thing clear: engineering delivered the risk report on one side, finance closed the balance sheet on the other, and risk transfer was renewed by a third party.
Each did its own part well. What was missing was someone looking at the space between the three. That was exactly where the risk lived — sitting on the balance sheet, waiting for a reading that pulled it all together.
Two sides of the same table.
Waldemir Queiroz
Thirty-five years across industrial operations, risk engineering, corporate insurance and finance leadership.
Led recoveries of severe incidents, structured transfer programmes, and sat on both sides of the table: the side that quantifies the risk and the side that decides how much capital to allocate.
Retention Engineering is a technical reading of how risk shows up on the balance sheet — and of how the company can see, measure and decide on it.
The first step is not to change the structure. It is to see the exposure.
We start with a short conversation, with no advance preparation. From what is already readable from outside about your operation, we show how the integrated reading of risk, capital and transfer works. If it makes sense to both sides, we go deeper after that.
Outside-in reading
A short conversation drawing on public information. No NDA, no preparation. Tests whether common ground exists.
Diagnosis
A map of the relevant exposure. An inventory of the three readings as currently practised. Where they fail to meet.
Modelling
PML by scenario. Aggregate AEL. RBC adjusted for capital, liquidity and covenants. Calculation of ERE.
Decision
Recommended architecture. What to retain, transfer, mitigate, and in what order. Premises, limits and review triggers, all recorded and auditable.
Four objects. One decision. The common language between operations, finance and risk transfer.
The method does not end in a report. It ends in four auditable objects that stay with the company and can be revisited at the next capital window.
Relevant Exposure Map
Which losses truly affect capital, cash, covenants and continuity. A materiality filter.
Effective Retained Exposure
How much remains on the balance sheet after effective recovery. A single number, directly comparable against RBC.
Prudential capacity and scale
RBC adjusted for capital, liquidity, covenant headroom, concentration and cash cycle. A four-band scale to locate the ERE.
Decision architecture
What to retain, what to transfer, what to mitigate, and what to review periodically. Premises, limits and triggers recorded.
The risk is already on the balance sheet. Retention Engineering turns that inheritance into a capital decision.
How much of the retention sitting on your balance sheet today was actually decided?
Start with the right question.
A 30-minute initial conversation is enough to test whether this reading makes sense for your operation. No advance preparation, no commitment, and no need to share confidential data.
The conversation draws only on public information and on a preliminary outside-in reading. The objective is not to issue a diagnosis. It is to test whether there is a relevant decision agenda.
Or forward this page to someone in finance or operations. The conversation adjusts to the audience.