The changes in accounting are potentially radical, but will they silence the critics? What problems does IFRS solve? And how will it affect insurers?

IFRS will provide a single basis of reporting for filing across borders. In addition to being mandated for commercial companies in many countries, it may soon be acceptable for insurance companies here in the United States and elsewhere. The value of this fact alone should not be underestimated, since a widely accepted standard aids both investors and companies needing capital.

Regulators are interested in IFRS. From their standpoint, it could form the basis for regulatory reporting at some point in the future, again potentially across borders. And, for better or for worse, taxable income may eventually be derived from IFRS. The benefits of a single basis of reporting for all filing purposes are lower costs for financial reporting and the opportunity to reduce complexity and to improve controls.

Uniform reporting will enhance comparability and transparency, but these attributes will be delayed for insurers. An interim standard for insurance contracts allows companies to continue existing accounting policies (with some modifications) until the International Accounting Standards Board can agree on the principles that should apply to insurance. While the IASB has set itself the goal of producing a draft Phase II standard by mid-2005, it appears that the earliest possible date for reporting under the final Phase II insurance contracts standard, which could be fair-value reporting, is 2008.

When the final standard arrives, will reporting be improved? The criteria the IASB itself uses to make that determination are relevance and reliability for the user.

Is IFRS more relevant?

The answer depends, of course, on the final standard. While asserting that it has not made up its mind, the IASB has devoted most of its resources to evaluating a fair-value measurement model for insurance liabilities. In IFRS, the “fair value” of a liability is the amount for which it could be settled in an arm's-length transaction.

Fair-value measures value insurance contracts using principles applied to other financial instruments, such as option-pricing techniques. Historically, fair-value measures have not been used for valuation of insurance liabilities, except in purchase situations where liabilities are sometimes adjusted or an intangible asset is recorded for the value of future profits. Africa has adopted a fair-value standard for insurance, and Canada uses a discounted cash-flow method for valuing liabilities that has many of the same dynamics as fair value. The UK and Australia also are familiar with such approaches.

In addition, some techniques used for risk measurement and capital management employ modeling approaches that have much in common with fair-value measures.

Many insurance companies report embedded values, typically by recognizing an intangible asset for the value of future profits. These have a fair-value flavor, but the IASB objects to embedded values on technical grounds, such as opposing the capitalization of investment margins.

While there may be some precedent for fair-value methods, there is no generally accepted practice. Furthermore, companies' processes for risk measurement are not designed to meet the demands of timely financial reporting.

IFRS is likely to be more volatile than most existing regimes, but proponents of IFRS note that volatility is a fact of life. Smoothing can mask risks and allow the “management” of results.

But does the volatility associated with IFRS reflect the nature of the business, or is it just the technical result of applying the accounting rules? There is real volatility in underwriting results and in the movement of asset values. Net “bottom-line” volatility can be reduced through closely matching assets and liabilities and valuing those on the same basis. However, induced volatility in profit or loss could easily appear under IFRS due to the inconsistent measurement of assets and liabilities, as could be the case under the interim rules for insurance.

A determination that IFRS is relevant often hinges on the belief that, for long-duration contracts in particular, short-term market movements are irrelevant, since companies are generally able to weather market volatility. Also, a conviction that fair-value measures are the key to assessing companies' ability to manage all the risks associated with their asset and liability portfolios supports the view that IFRS is relevant.

Is IFRS more reliable?

Insurance products are complicated, and the accounting rules are complex. Results from one company to the next have potential differences in approaches and in assumptions. Results also are sensitive to assumptions, and so movements in liabilities may reflect changing views about the future, not just the events of the period. To what extent should changing variables related to the future affect todays performance report?

Even more fundamental is the concern that valuing contracts by market-consistent pricing techniques may not even be valid. Market-consistent pricing assumes that markets are efficient and that frictional costs have a minimal impact. These points are generally not true for insurance contracts, which makes calibrating models difficult.

Despite these concerns, there is no need to be cynical about the direction of IFRS and the potential it offers for greatly improved financial reporting. Actuaries are up to the task of developing disciplined approaches, reaching agreement on methods, and implementing controls. Therefore, fair-value measurement, when combined with appropriate disclosures, can be relevant and reliable.

Moreover, acknowledging the link between appraisals and the fair-value measures, the market must have some confidence in insurers ability to use fair-value measures, judging from the amount of money changing hands in recent life insurance mergers and acquisitions.

A reporting regime that is consistent with an economic view of insurance will support the industry trend toward improving the understanding of risk and will align reports to shareholders with good capital management practices.

The issue, then, is whether the IASB will settle on an approach to insurance accounting that meets its objectives. Because the IASB currently is on course to define new practices rather than to codify existing practices, some believe it would be a mistake to go forward without a field test.

The stakes are high. Whatever decision the IASB ultimately reaches, the impact on insurance companies will be far-reaching. Some of the biggest effects will be felt in the areas of:

Product design and pricing: Depending on market acceptance, companies will re-examine and revise product features, particularly asset-intensive products and those with guarantees and options that have the greatest influence on fair values.

Reinsurance programs: Insurers may shift from accounting/regulatory arbitrage to hedging and risk-sharing arrangements.

Communications to shareholders: Insurers will have to totally revamp their communications to explain results under the new accounting principles.

Compensation and incentive arrangements: Often tied to reported profits, compensation arrangements will need to be revisited in light of new profit measures.

Much depends on the IASBs deliberations and conclusions regarding the appropriate accounting for insurance. In the meantime, there is uncertainty about what direction the IASB will take. The IASB and insurers will need to work closely together to reach a consensus, such as by conducting a pilot test in the field. Insurance accounting is certain to undergo significant change, and the industry will adjust. Despite some potential drawbacks, IFRS may offer real benefits in the longer term, especially if it results in a closer alignment of financial reporting and capital management.

Jim Milholland is a partner in Ernst & Youngs Insurance and Actuarial Services practice. He is currently based in the firms London office.

Reproduced from National Underwriter Edition, April 23, 2004. Copyright 2004 by The National Underwriter Company in the serial publication. All rights reserved.Copyright in this article as an independent work may be held by the author.