Provisions Are Predictions: Why Caribbean Boards Must Govern Reserves Before the Audit

This is the third of four papers in The Caribbean Actuarial Imperative, an editorial series from Dawgen Global — an independent, integrated, multidisciplinary professional services firm spanning eleven service disciplines across more than fifteen Caribbean territories. The first paper introduced the Long‑Horizon Test for identifying the board decisions whose financial consequences extend beyond the quarterly window. The second examined pension valuation and the questions a trustee is implicitly answering by approving it. This paper turns to the domain where the discipline of those first two papers becomes an audit question: the reserving of liabilities whose ultimate cost is uncertain and whose settlement is years away.

It is written for the people who sit on both sides of the reserve: the directors and audit committee members who approve the financial statements that contain it, the chief financial officers who book it, the appointed actuaries who estimate it, and the external auditors who must form an opinion on whether it is fairly stated. Reserving is the one long‑horizon obligation that a Caribbean institution cannot quietly leave un‑examined, because the auditor is required to examine it every year — and the quality of that examination depends on whether anyone has governed the estimate before it reaches the audit file.

Most numbers in a set of financial statements record something that has happened. A reserve records something that will happen, on the basis of something that already has — and the distance between those two events is exactly where reserving discipline lives, and most often fails.

1.   The number that is a forecast

When a finance company books revenue, it records a transaction that has occurred. When it records a cash balance, it states a fact that can be confirmed. When it recognises a reserve for claims, by contrast, it records an estimate of payments it expects to make in the future for events that have already taken place — a road accident already on the books of a motor insurer, a workplace injury already suffered, a professional error already committed but not yet alleged. The cost of the event is fixed in principle. It is simply not yet known.

This is the defining feature of a reserve, and the source of every difficulty that attends it. The reserve is a number on the balance sheet that looks exactly like the cash balance beside it — a single figure, audited, signed. But it is a forecast wearing the clothing of a fact. The principle the first two papers established holds here in its most acute form: the number is the output of a methodology, and the methodology rests on assumptions. When the assumptions are sound, the reserve is sound. When they are stale, optimistic, or borrowed from a book of business that does not resemble the institution’s own, the reserve is a misstatement that no amount of arithmetic accuracy will reveal.

2.   Why the long tail is the hard part

Reserving is difficult in proportion to the length of the tail — the time between the event that creates the liability and the final settlement of it. For some obligations the tail is short. A property claim from a storm is usually reported within days and settled within months; the estimate is bounded and corrects quickly. For others the tail is long, and it is the long‑tail liabilities that defeat un‑actuarial reserving.

Consider the categories a Caribbean institution most commonly carries. Liability and professional‑indemnity claims may be reported years after the act that caused them and litigated for years after that. Healthcare and disability claims develop as conditions evolve. Warranty obligations on a product sold this quarter generate a claims pattern that only stabilises after three years of data. Each of these shares the feature that makes reserving an actuarial discipline rather than a clerical one: a large part of the ultimate cost is not yet visible in the institution’s own records at the moment the reserve must be struck.

That invisible portion has a name — incurred but not reported, the claims that exist in the world but have not yet arrived on the institution’s desk. Estimating it is not a matter of adding up known claims and rounding upward. It is a matter of inferring, from the pattern of how past claims have developed, how much of the present year’s ultimate cost is still to emerge. That inference is the core of reserving actuarial work, and it is precisely the part that a reserve booked on instinct, or on last year’s ratio, gets wrong.

3.   Where the auditor meets the actuary

Reserving is the domain in which financial‑statement assurance and actuarial method are not merely adjacent but interdependent. The external auditor is required to obtain sufficient appropriate evidence that the reserve is fairly stated. But the reserve is the output of an actuarial estimate, and the auditor is rarely an actuary. So the auditor must do one of three things: rely on a management’s expert, engage an auditor’s expert, or challenge the estimate using audit techniques that were not designed for forecasts. Each route has a failure mode, and in the Caribbean each failure mode is observed.

Where the auditor relies on the appointed actuary as management’s expert, the assurance is only as strong as the auditor’s evaluation of that expert’s competence, objectivity, and methodology — an evaluation that is demanding to perform well and easy to perform in form only. Where the auditor engages an independent actuarial specialist, the assurance is stronger but the cost and the coordination are higher, and smaller engagements often cannot bear them. Where the auditor proceeds without specialist input, testing the reserve against prior years and management’s representations, the audit can confirm that the number was calculated consistently while remaining silent on whether the basis was ever right.

The constructive position is not to assign blame among these routes but to recognise what they have in common: each works far better when the reserve has already been governed before it reaches the audit. An estimate whose assumptions were made explicit, defended, and tested for sensitivity at the point it was struck is an estimate the auditor can evaluate efficiently. An estimate that arrives as a single figure with its reasoning buried is one the audit must either reconstruct at cost or accept on trust. Reserving discipline, in other words, is upstream of audit quality — and is the institution’s responsibility, not the auditor’s.

4.   The Dawgen Global Reserve Integrity Questions

The seven questions below are the framework Dawgen Global uses to govern a reserve before it is booked and before it is audited. They are the reserving counterpart to the Valuation Governance Questions of the second paper, and they are addressed to the audit committee and the finance function rather than to the actuary alone. A reserve whose owners can answer all seven is one the external auditor can assure efficiently and the board can stand behind. A reserve where the questions produce silence has located, early and cheaply, the exposure that a regulatory examination or a transaction would otherwise locate late and expensively.

• What method was used to estimate the ultimate cost — and how much of the reserve represents claims already reported versus claims incurred but not yet reported?
• What development pattern was assumed, whose experience does it rest on, and does that experience resemble this institution’s own book rather than a borrowed one?
• How has the reserve for prior years developed against what was originally booked — has the institution historically under‑reserved, over‑reserved, or estimated well?
• What is the range of reasonable outcomes around the booked figure, and where in that range does the booked figure sit — prudent, central, or optimistic?
• Which single assumption would move the reserve most if it proved wrong, and what is the financial consequence if it does?
• Is the basis consistent with last year’s, so that the movement can be decomposed into changes in assumption, changes in claims experience, and changes in the underlying exposure — or does it arrive as one unexplained figure?
• If the external auditor, a regulator, or an acquirer tested this reserve tomorrow, would the basis withstand the scrutiny — and if not, what would we change before they do?

An audit committee that asks these seven has not displaced the actuary or the auditor. It has done the one thing that makes both their work reliable — it has governed the estimate rather than ratified it.

5.   Where Caribbean reserving most often breaks

Across the engagements Dawgen Global runs in the region, the weaknesses in reserving recur with enough regularity to be named. They are structural and addressable, and they cluster well beyond the insurance carriers that are usually assumed to be their only home.

• The borrowed development pattern. A reserve estimated using claims‑development factors drawn from another market, another book, or a software default — rather than from the institution’s own claims history — is a recurring source of both over‑ and under‑statement. The pattern of how claims emerge is specific to the book; importing it wholesale imports its errors.
• The incurred‑but‑not‑reported blind spot. Institutions that reserve only for the claims they can see, with no rigorous estimate of those they cannot yet see, systematically under‑state long‑tail liabilities — and discover the gap only as the unreported claims arrive in later years, against a reserve that was never set aside for them.
• The reserve set to the result. Where the reserve is adjusted, consciously or not, to produce a desired profit or capital outcome rather than struck on its own merits, the estimate ceases to be an estimate. This is the failure that regulatory examinations and acquirers look for first, and it is the most damaging to find late.
• The non‑insurance reserve no one treats as actuarial. Credit‑loss provisions in credit unions and co‑operatives, warranty reserves in manufacturers and distributors, self‑funded medical‑scheme liabilities in employers, and litigation and contingent‑liability provisions in any institution are all long‑tail estimates of uncertain ultimate cost — and are routinely booked without the discipline that the word ‘reserve’ should trigger, because they do not sit inside an insurance company.

6.   Beyond insurance: the reserves hiding in plain sight

The most consequential point in this paper for the typical Caribbean board is that reserving discipline is not an insurance topic. The institutions that carry the least‑governed long‑tail reserves are frequently not insurers at all. A credit union’s provision for expected credit losses is a reserve in every meaningful sense — an estimate of future losses on events, namely the loans, that have already occurred. A distributor’s warranty provision is a reserve. An employer’s self‑funded health plan carries a reserve for claims incurred but not reported exactly as an insurer does. A company facing litigation carries a contingent‑liability estimate that is a reserve under another name.

Each of these is audited. Each is approved by a board. And each is, far more often than in the insurance sector where the discipline is at least expected, booked without an explicit method, an own‑book development assumption, or a sensitivity range. The Reserve Integrity Questions apply to all of them without modification. An institution that has never thought of itself as carrying actuarial liabilities is frequently the institution carrying the most poorly governed ones.

7.   What reserving‑grade actuarial advisory looks like

The actuarial work that produces durable assurance value in reserving shares the character the series has identified throughout, sharpened to the audit interface.

• It is scoped to support the audit, not to duplicate it — producing the explicit method, own‑book development assumptions, and sensitivity range that let the external auditor form an opinion efficiently rather than reconstruct the estimate from scratch.
• It rests the development pattern on the institution’s own claims experience wherever credible data exists, and is explicit about the adjustments made where it does not.
• It separates and quantifies the incurred‑but‑not‑reported component rather than folding it invisibly into a single figure.
• It reports the range of reasonable outcomes and states where in that range the booked reserve sits, so the board governs a prudence decision rather than ratifying a point estimate.
• It is integrated with the audit, the capital and regulatory position, and — for non‑insurance reserves — the accounting policy under which the provision is recognised, so the estimate is sound on every dimension on which it will be tested.

8.   Where this series goes from here

This is the third of four papers in The Caribbean Actuarial Imperative. The closing paper turns from the obligations already on the balance sheet to the one whose measurement is only now being demanded:

• The fourth paper will examine climate risk quantification — the forward frontier of Caribbean actuarial work, where the long‑horizon reasoning of the first three papers is applied to exposures that ratings agencies, correspondent banks, and supervisors are beginning to require institutions to measure, and where regional capability will most clearly differentiate the leaders over the next five years.

Each paper follows the same discipline as this one: written for Caribbean boards, not for actuarial specialists; technically rigorous but accessible; vendor‑neutral and independent; and grounded in the engagements Dawgen Global runs every week with clients across the region.

9.   How Dawgen Global delivers actuarial advisory

Dawgen Global is an independent, integrated, multidisciplinary professional services firm. Its actuarial work is delivered through the firm’s Actuarial Services Division, which operates within the broader Risk Advisory practice and is led by senior licensed professionals with internationally recognised credentials. The Division’s work covers pension valuation and scheme governance, reserving methodology and long‑tail liability estimation, capital adequacy and stress testing, climate risk quantification, and the application of actuarial methods to non‑insurance domains. It operates alongside the firm’s ten other service disciplines — Audit & Assurance, Tax Advisory, IT & Digital Transformation, Risk Management, Cybersecurity, HR Advisory, M&A Advisory, Corporate Recovery, Business Advisory & Strategy, Accounting BPO and Virtual CFO services, and Legal Process Outsourcing.

The integration is, in reserving, the entire point. A reserve that needs an actuarial estimate also needs an audit that can rely on it, an accounting policy that recognises it correctly, a regulatory position that defends it, and — in a transaction — a due‑diligence view that prices it. An independent, integrated, multidisciplinary firm is the natural home for that combination. The structural separation between the independent advisor and the technology partner that the firm argues for across its work applies to actuarial modelling as well: the reserving software and data infrastructure are selected with each client, downstream of the methodological judgements this paper is about.

ABOUT THE ACTUARIAL SERVICES DIVISION

The Dawgen Global Actuarial Services Division operates within the firm’s Risk Advisory practice and is led by senior licensed professionals with internationally recognised actuarial credentials. The Division’s work covers pension valuation and scheme governance, reserving methodology and long‑tail liability estimation, capital adequacy and stress testing, climate risk quantification, and the application of actuarial methods to non‑insurance domains including credit unions and co‑operative societies, self‑funded employer benefit schemes, contingent commercial obligations, and merger and acquisition due diligence.

Dawgen Global is an independent, integrated, multidisciplinary professional services firm headquartered at 47 Trinidad Terrace, New Kingston, Jamaica, operating across more than fifteen Caribbean territories. The firm spans eleven service disciplines and works with a curated network of global technology partners and vendors to design, develop, and implement effective client solutions. Dawgen Global is independent and integrated — not affiliated with, a member of, or backed by any external network.

dawgen.global   •   [email protected]   •   47 Trinidad Terrace, New Kingston, Jamaica

This thought leadership paper is published for general information only and does not constitute legal, tax, audit, actuarial, regulatory, or investment advice. The Reserve Integrity Questions described above are illustrative of Dawgen Global’s advisory methodology and should be adapted to the specific circumstances, regulatory environment, and risk profile of each institution. Independent professional advice should be obtained before acting on any matter discussed in this paper. © 2026 Dawgen Global. All rights reserved.

 

About Dawgen Global