2 Insurance Companies

Howell Zee
Published Date:
April 2004
  • ShareShare
Show Summary Details
Robin Oliver

Insurers are important financial intermediaries. They play a significant role in the financial system as vehicles for savings (policy premiums are not used immediately to meet claims; they are instead invested, earning policyholders a return when they either collect on claims or have their premiums reduced) and risk pooling (insurers pool the risks of different policyholders, so that those whose insured risk eventuates are compensated in part out of premiums paid by those who do not file claims). The insurance office itself provides an intermediary service. It underwrites some of the residual risk of the risk pooling and may assume some of the savings risk. An example of the latter is a traditional, nonparticipating annuity where the insurer guarantees the annuity payment irrespective of the investment performance of the fund from which the annuity is financed.

Types of Insurance

Life insurance broadly encompasses insurance contingent on human life. Standard life insurance policies provide for monetary payments to an insured individual’s designated beneficiaries on his or her death. Other forms of life insurance include annuities and income maintenance insurance, payments of both of which cease when the insured dies. Life insurance tends to have a high proportion of a savings element because most policies have multiyear terms and allow payoffs to the policyholders during their lifetime. This means that the premiums are used to build up an investment to finance the policy. However, the extent to which life insurance incorporates a savings element varies according to the type of policy. Some main policy types include term (or annual), whole life (flat annual premiums set for a number of years), endowment (the policy is paid out when the insured reaches a set age), unbundled (when the savings element is separately identified), and annuities. Policies also vary in that they may be participating (the policyholder bears some of the savings risk) or non-participating (the savings risk is assumed by the life office).

General insurance—also known as catastrophe and property insurance—involves the insurer’s assuming or underwriting losses from events not contingent on human life. Such insurance has a greater risk-pooling but smaller savings element than life insurance because the contracts generally have shorter terms. It covers losses caused by a broad range of events, including earthquakes, wars, accidents, ill health, and even low profits for any reason.

Reinsurance is a transaction between two insurers: one purchases insurance from the other to cover part or all of the risks that the purchasing insurer does not wish to carry in full. Reinsurance varies in nature (depending on the risk reinsured) but tends to be mainly for risk pooling.

Insurers are not the only financial intermediaries that can pool and bear risk. For example, foreign exchange contracts that cover foreign exchange risks are also vehicles of risk pooling. As a result, the boundary between insurance and other financial intermediaries can be blurred—and increasingly so as the regulations demarcating different aspects of the financial sector have been reduced in most countries. As other forms of financial intermediation have been able to take on more characteristics of insurance, insurance has been able to take on more characteristics of other forms of financial intermediation. As examples, a normal coupon-paying bond can be written in the form of an annuity, and a normal financial investment, in the form of a general insurance policy. This increasing substitutability of insurance and other forms of financial intermediation increases the importance of having tax rules that treat all segments of the financial sector in a neutral way.

Organization and Regulation

Life and general insurers have two basic organizational forms: mutual and proprietary. Mutual insurers are owned by their policyholders, who are the ultimate bearers of savings and residual risks. In contrast, owners of proprietary insurers are separate from the policyholders, and, as a result, the residual—but not, in general, the savings—risk is borne by the insurers. Proprietary companies may be domestically or foreign owned.

The international trend has been for mutual companies to demutualize and become proprietary and for proprietary insurers to operate internationally (often through foreign ownership). Demutualization seems to be driven by the need for better management accountability and the efficiencies that come from that. The internationalization of insurance is the product not only of general economic globalization but also of the relaxation of regulations requiring insurers to be residents of the country in which they write policies. The internationalization of the insurance market is now well developed. According to data from the Organization for Economic Cooperation and Development (OECD), the foreign company share of nonlife insurance in 1997 was 37.3 percent in the United Kingdom, 12.7 percent in Germany, 8.8 percent in the United States, and 3.6 percent in Japan.1

Reinsurers tend to be proprietary and international.

Most countries have regulatory regimes for different aspects of the financial sector, to both protect it from the collapse of one of its constituent parts and reduce the risk of default. Because of their particular features, general insurance and life insurance usually have regulatory regimes specific to them and different from those that apply to other financial intermediaries (such as banks). Among the regulations governing the insurance sector are accounting requirements that insurers adopt a conservative approach to their financial reporting. For example, insurers are sometimes required to provide reserves for future contingencies.

Regulations often restrict general and life insurers from offering noninsurance-related financial intermediation services. The aim is to prevent a collapse in one area of the financial system from spreading to other areas. However, the international trend has been to move away from regulated segregation of the financial sector.

Tax Policy Objectives

Tax rules for the insurance sector should, as far as possible, achieve the following objectives:

  • Neutrality between different forms of financial intermediation. It is especially important to have consistent tax rules for easily substitutable activities. Otherwise, financial activities will be encouraged to move from industries with high taxes to those with lower taxes, irrespective of the relative efficiency of the industries. In highly regulated financial industries, some financial products may not be close substitutes, thus reducing the distortions caused by different tax rules as applied to them. However, as noted above, insurance and financial intermediation are becoming more substitutable. While, ideally, tax rules should be neutral between different industries in the financial sector, they also need to accommodate the special features or requirements of different sectors. Thus, in both general and life insurance, tax rules need to get at an appropriate measure of the income of the insurance office after taking into account the deductibility of premiums and the tax treatment of claims in the hands of policyholders.
  • Consistency with broad, overall government policies. If retirement savings and pensions are subject to special tax rules (such as preferential tax treatments), life insurance rules should be similarly formulated. Likewise, if the corporate income tax (CIT) integrates the tax position of the company and the shareholder (by, for example, an imputation system) to avoid overtaxing equity income, insurance tax rules should do the same. In addition, if the personal income tax on savings income is subject to variable rates or is exempt, then this should be reflected in the extent to which tax is levied on the savings component of the insurance policy.
  • Avoidance of conflict with the regulatory regime. For example, tax rules should not encourage the use of accounting approaches inconsistent with those required by the regulations. As another example, the regulatory regime could be undermined if the tax rules encourage mutual insurers to reduce their taxable profits by returning income tax free to policyholders in the form of lower premiums. This could also undermine the solvency of those insurers. The regulatory regime can be used to reinforce the tax regime and vice versa.While not contradicting the regulatory rules, tax rules may be different because the two sets of rules serve different purposes. Regulatory rules are concerned with the overall solvency of financial entities and the financial sector. In contrast, the objective of income tax rules is to measure the annual income of financial entities accurately. Thus, for example, simply because regulatory rules require insurers to provide for future contingencies in their financial reports does not mean that tax rules should necessarily allow those contingencies to be deductible against income.
  • Appropriate tax treatment of the profits attributable to insurers’ owners (from financial intermediation and risk-taking services) as well as policy-holders’ income from savings.
  • Administrative compatibility with existing tax procedures to minimize enforcement and compliance costs. For example, if individuals are not required to file tax returns, this is likely to inhibit taxing life policyholders directly on a share of life office profits attributable to such policyholders. The complexity of the tax rules adopted also needs to take into account the technical expertise available to insurers in complying with the rules and to tax authorities in administering them.
  • International consistency. It is necessary that the tax rules adopted can be applied in an international context, so that appropriate tax can be levied on foreign owners of insurers operating domestically, on domestic owners of insurers operating overseas, and on domestic holders of insurance policies issued by foreign companies. In addition, if the tax treatment does not qualify as income tax (because, for example, the tax is imposed on a non-income base), the protection provided against double taxation by double tax agreements could be lost, which, in turn, could hinder growing international insurance linkages.

Life Insurance

Under normal CIT rules, premiums and investment earnings would be taxable income, and claims would be deductible expenses, but deductions are not usually allowed for reserves and provisions until the expenses to which they relate are reasonably certain, provable, quantifiable, and legally binding. If such rules were applied to life insurance, however, the result would be inaccurate annual measures of what is, in substance, the income generated by the life office on behalf of policyholders and owners, for the following reasons:

  • The long-term nature of most life policies makes it difficult, using normal tax rules, to match expenses to the income to which they relate for a particular year. Taxing premiums as income without a deduction for a provision for future claims would be the equivalent of taxing a bank on the deposits it receives. This is less of an issue for short-term policies because claims arise in the same year or soon after premiums are recognized, providing an approximate match of income and expenditure.
  • The mix of savings return, savings and risk intermediation, and risk pooling (all of which can give rise to income) is inherent in a life policy. Under a neutral income tax regime, the savings return should be taxed on the same basis as the return to a bank deposit. The intermediation return is the life office’s profit and should be taxable as such, while the risk pooling between policyholders should be free of tax, since it represents a transfer of wealth, not the direct generation of income. With life insurance, however, it is often difficult to identify these different returns separately.
  • The ownership of the various elements of life office profits is divided between policyholders and proprietors in the case of proprietary companies, and between policyholders in their different capacities as owners and customers in the case of mutual companies. Again, the proper attribution of overall profits may be difficult to determine on a periodic basis as required under normal income tax rules.

These features of life insurance are common to most countries. Combined, they appear to prevent the simple application of normal income tax rules to life offices. As a result, countries generally have specific income tax rules that apply to life offices, but adopted approaches vary considerably between countries. In most cases, individual country rules tend to be very detailed and subject to constant change. This is indicative of the significant challenges life insurance poses for tax policymakers.

To clarify the problem, it is necessary to separate out the two distinct forms of income inherent in the activities of a life office.

Underwriting Income

This is the business income of the life office. It may be attributable to proprietors or to policyholders, depending on the nature of the policies and whether the office is a mutual or not. To avoid tax rules that favor or penalize the insurance industry, it would be desirable for life offices to be taxed on this income under normal income tax rules—that is, profits should ideally be calculated as revenue less expenses. Life office revenues are premiums (P) and investment income (I), while expenses are claims (C) and all other outlays (E) related to underwriting and investment. Profits are therefore: (P + I)-(C + E). However, as previously noted, taxing a life office in this way would not produce a good reflection of its profits. Because of the long-term nature and other features of life insurance, reserves must be taken into account to match (on an annual basis) revenue and expenses.

As a result, first, a reserve (the actuarial reserve) needs to be added to claim deductions to reflect the accrual of claims over time.2 The present value of the liability for claims rises as the expected date of death of the insured individual (according to mortality tables) draws nearer. The present value of claims also rises as past investment earnings and discounted future investment earnings backing the policy rise.

Second, a reserve needs to be deducted from underwriting expenses to account for the fact that many such expenses (for example, agent commissions) are front-end loaded but are incurred to produce a long stream of underwriting income. Expense reserves are often netted off against the actuarial reserve to give the standard underwriting income formula of: [(P + I) - (C + E) - V], where V is the change (usually an increase) in net reserves. In addition, V needs to include discontinuance profits or losses from surrendered policies (or the formula has to be adjusted to take such profits or losses into account).

An important feature of the formula is that, although it includes investment income, such income is largely netted out in the reserves that rise with investment earnings. Under most policies, investment earnings accrue to the policy-holder and not to the life office. The underwriting profit therefore should include investment earnings only from nonparticipating policies in excess of the estimated earnings in accordance with which the policy was written. The difficulty of applying this formula for tax purposes is that it is subject to many judgments and uncertainties in reaching an appropriate reserve figure. Factors that determine reserve levels include (1) the method for spreading underwriting expenses, (2) the discount rate to be used in valuing expenses and investment earnings, (3) the level of future earnings backing policies, and (4) the appropriate mortality rates to be used.

When tax rules attempt to tax underwriting profits, there is a tendency to rely on regulatory or accounting standards. However, since these focus on the solvency of the life office, they tend to be conservative and require generous reserves. As an alternative, the United States and some other countries have set out specific, detailed policy rules for reserves for tax purposes. In addition to being complex, these rules do not necessarily improve the result. The United States rules are generally regarded as being concessional.

The difficulties in calculating life office reserves are compounded by the small margins typical of the insurance industry (both life and nonlife). Thus, small variations in reserve calculations can have a significant impact on reported underwriting profits.3 International experience has been that taxing life offices on the basis of actuarial reserves produces little or no taxable income.

For policies written by foreign companies or across international borders, it is also necessary to ensure that income and expenses are appropriately allocated to different tax jurisdictions. Expenses directly related to the writing of a policy (such as commissions) are generally relatively easy to allocate. Overhead expenses and investment income and expenses can be more difficult. In such cases, even transfer pricing rules (such as the OECD arm’s-length principle) can be difficult to apply. Savings income of the life office should be attributed to the policies it backs to the extent that offshore savings income is to be taxable (see below). Such income should also be included in reserves for the purpose of calculating underwriting income.

Savings Income

This is the income derived by the life office in its role as a savings intermediary, and it is ultimately attributable to policyholders who are substantially in the same position as bank depositors or mutual fund investors. This savings income is investment income minus investment expenses, or (I - E). As previously noted, this is excluded from underwriting income by the reserve formula.

Savings income can be measured each year as the net investment income of the life office that is not attributable to proprietors. This ensures that such income is fully taxable as it accrues. However, achieving this result is problematic. There is a need to measure the extent to which investment income is attributable to policyholders as opposed to proprietors. There is also the problem of taxing income in the hands of policyholders at their applicable tax rates. Finally, there is the question of whether the net investment income—when taxable—should be taxed under ordinary tax rules or under rules specific to the insurance business.

The policy objective of neutrality between different forms of financial intermediation would suggest that ordinary tax rules should apply. However, depending on what those ordinary tax rules are and the insurance rules adopted, measuring net investment income on the basis of ordinary tax rules could be highly favorable to insurers. In other words, ordinary tax rules for measuring an insurer’s net investment income could distort financial markets in favor of insurers. That is because, for practical or policy reasons, ordinary tax rules do not tax all economic income comprehensively. Capital gains on debt and equity instruments may not be taxable. Even if all such gains are taxable, it is unlikely that ordinary tax rules will always tax gains on a full accrual basis. The under-taxation of insurers that would result from ordinary tax rules underestimating net investment income would not of itself distort neutrality between different industries of the financial sector. However, insurance tax rules may allow a deduction for claims and claim reserves financed from untaxed investment income. In this case, insurers receive a deduction for costs that are matched by untaxed income, leaving them in a favorable position relative to other financial intermediaries. Determining the appropriate rules for measuring net investment income in the insurance tax base therefore requires consideration of the details of ordinary investment tax rules and the specifics of the insurance tax rules. The general rule in OECD countries is that life offices are taxed on realized gains on securities. An exception is Belgium, which exempts capital gains on shares.4

An alternative approach is to tax income as it is distributed to, or applied for, the benefit of policyholders. Savings income can be applied to the benefit of policyholders in a number of ways:

  • An increase in the cash surrender value of the policy as notified to policy-holders. This is not a cash distribution to policyholders but the amount for which the policy can be turned into cash. Relative to taxing income as it accrues to the life office, there is a deferral advantage, but this deferral advantage is smaller than that based on cash distribution.
  • Policy dividends. These are cash payments to continuing policyholders in cases where the investments backing the policy exceed the amount necessary to fund future benefits. Taxing savings income on the basis of policy dividends provides a substantial deferral advantage.
  • Increases in coverage. Investment earnings can be used by the life office to increase policy coverage or to reduce the costs of existing coverage.
  • Policy loans on favorable terms such as low rates of interest. The benefit to the policyholder is the difference between market and concessional terms.
  • Cash payments on surrender or maturity of the policy (including payments on death) whether in a periodic form (a pension or annuity) or as a lump sum.

The following are problems with taxing policyholder benefits:

  • Tax on savings income is deferred until income is attributed to policy-holders. This provides a deferral advantage and can encourage life offices to structure their policies so as to maximize that advantage.
  • Unless premiums are deductible in the hands of policyholders (other than as a specific tax concession), the taxable benefits of policyholders should exclude the return of premiums and claims to the extent to which they represent premiums paid by other policyholders because of earlier-than-expected deaths. These calculations would be extremely complex. If premiums are deductible, then investment income at the life office plus benefits provided to policyholders should both be taxable to preserve neutral tax treatment with other forms of savings.
  • The result of taxing policyholder benefits is likely to be complex for policy-holders and can result in taxpayers facing a tax liability without receiving cash payments.
  • There are many ways life offices can allocate benefits to policyholders and it is likely to be difficult to ensure that tax is payable in all cases.

If policyholders are taxed on savings income, then it is necessary to ensure that they cannot escape taxation by using foreign life offices. There is also a need to ensure that life offices cannot avoid tax on domestic policies by writing policies offshore or by transferring income offshore by way of reinsurance (see below).

International Practices

The ideal would be to tax life offices on underwriting income (with nonconcessional policies for reserves) and policyholders fully on savings income as it accrues to their benefit. Since this is not practical, actual international practice varies between countries. There seem to be four broad approaches or models (Table 2.1).

Table 2.1.Tax Treatment of Life Insurance in Selected OECD Countries1
CountryUnderwriting income2Reserve policy3Policyholder tax4
Australia5ENot applicableTL
BelgiumENot applicableE
CanadaTRTL or TP
IrelandENot applicableTL
New ZealandTRTL
SwedenENot applicableTL
United KingdomENot applicableTL
United StatesTSTP
Sources: OECD, Taxing Insurance Companies (Paris, 2001); and author’s compilation.

The tax system of each country is subject to detailed rules that make generalizations difficult. For example, tax rules can vary depending on the type of policy involved. This table is therefore only a guide to different approaches that are adopted for the taxation of life insurance.

E means that underwriting income is exempt; T means that underwriting income is taxable.

R means that underwriting income is calculated using regulatory or accounting reserve requirements; S means that underwriting income is calculated using reserve requirements specific to income tax law.

E means that the savings component attributable to policyholders is not taxable in the hands of either the life office or the policyholder in some cases because, although benefits may be taxable, this is offset by a deduction for premiums; E+ means that the savings component is not taxable, but, in addition, policyholders can deduct premium payments from other income; TL means that the savings component is taxable at the life office level; and TP means that the savings component is taxable in the hands of the policyholders.

Australia is proposing to move to a T/R/E approach for nonparticipating or risk-based policies and an E/TP approach for participating or non-risk policies.

Sources: OECD, Taxing Insurance Companies (Paris, 2001); and author’s compilation.

The tax system of each country is subject to detailed rules that make generalizations difficult. For example, tax rules can vary depending on the type of policy involved. This table is therefore only a guide to different approaches that are adopted for the taxation of life insurance.

E means that underwriting income is exempt; T means that underwriting income is taxable.

R means that underwriting income is calculated using regulatory or accounting reserve requirements; S means that underwriting income is calculated using reserve requirements specific to income tax law.

E means that the savings component attributable to policyholders is not taxable in the hands of either the life office or the policyholder in some cases because, although benefits may be taxable, this is offset by a deduction for premiums; E+ means that the savings component is not taxable, but, in addition, policyholders can deduct premium payments from other income; TL means that the savings component is taxable at the life office level; and TP means that the savings component is taxable in the hands of the policyholders.

Australia is proposing to move to a T/R/E approach for nonparticipating or risk-based policies and an E/TP approach for participating or non-risk policies.

U.K. model—life office taxed on net investment income as proxy for policy-holder tax on savings income. This approach recognizes the difficulty of accurately taxing underwriting income by excluding it from the tax base. The life office is taxed on net investment income (I - E) as if it were just a savings intermediary paying tax on behalf of policyholders. In broad terms, this model is also followed by Australia (although Australia is in the process of changing its rules), Belgium (although Belgium uses cash surrender value rather than investment income), Ireland, and Sweden. The obvious disadvantage of this approach is that it leaves the pure insurance component preferentially taxed, and there can be complexities over issues such as the allocation of overhead expenses between investment and underwriting activities.

U.S. model—underwriting income taxed and policyholder benefits taxed on deferral. This model is also followed by Germany and, broadly, by Finland. It is also the model traditionally adopted by Canada, although Canada has now moved closer to the New Zealand model (described below). Underwriting income is taxed, and, thus, insurance activity is within the tax net, but the taxation of the savings component is deferred until savings income is attributed to policyholders.5 This gives life insurance a tax advantage in the savings industry.

Continental Europe model—concession applied to savings. This is the dominant model of continental Europe. Underwriting income is taxed, but either the savings component is exempt (Mexico and the Netherlands) or policyholder benefits are taxable (as in the U.S. model) but after being offset by premiums (Austria, Germany, and Spain). Luxembourg and Portugal allow premiums to be deducted and also exempt the savings component. They thus tax the insurance activity but deliberately provide a tax concession for life insurance savings.

New Zealand model—underwriting income taxed and policyholder benefits taxed on accrual. This is arguably the purest model, in that both underwriting and savings income is taxable on an accrual basis. However, as with the U.K. model, savings income is taxed at the life office level without regard for the varying tax rates applicable to individual policyholders. Canada has moved closer to this model (from the U.S. model), and Australia seems to be in the process of doing the same (from the U.K. model).

Because of the difficulties of taxing life offices, some countries have buttressed their normal income tax rules in this area with special rules. Canada, for instance, sets a minimal tax based on the capital of the life office, and Mexico has a general tax based on the life office’s assets. Some countries also levy transaction taxes on premiums. The difficulty with taxing premiums is that it does not take into account the different risks and income associated with different types of life insurance products. Hence, a premium tax cannot be considered an accurate substitute for tax on underwriting or savings income.

General Insurance

As with life insurance, there are two components to the profits of a general insurer: underwriting income, which equals premiums minus underwriting costs less claims, and savings (net investment) income. The issues are less complex for general insurance than for life insurance, but it would still not be unusual for an insurer to make a consistent underwriting loss while making an overall profit if investment income is taken into account. The issues for savings income are the same as for life offices. Underwriting income is considered below.

Underwriting Income

It is easier to measure the underwriting income of general insurers than of life insurers, and, generally, less deviation from normal tax rules is required. That is because general insurance tends to be shorter term, with only a minor savings element. Hence, taxable profits can be measured more easily as the difference between premiums plus investment income less claim expenses. The inherent risk aspects of insurance and the fact that insurers will want to spread risks over time and geographically, however, still make it desirable to alter these basic calculations by reserve and provisions adjustments that would not normally be allowed under general tax rules that eschew adjustments for uncertainty.

As a result, various reserve and provisions adjustments can be found in international practices. Most countries allow general insurers a deduction for the unearned premium reserve and outstanding claim reserves. The unearned premium reserve, normally calculated on a daily basis, is the portion of gross premium that relates to the covering of risks beyond the insurer’s balance sheet date. The outstanding claim reserves consist of the loss reserve, an actuarial estimate of losses reported but not yet paid out, and the incurred but not reported reserve (generally known as the IBNR), an actuarial estimate of losses that have occurred during the year but have not been reported. An example would be a loss from an asbestos-related disease. The insurer’s liability arises at the time the disease is contracted, but the disease may not become evident and, thus, may not be reported until many years later. It is generally accepted that a deduction for such reserves would provide a better reflection of underwriting income than not allowing the deduction, although it should be balanced by rules that defer prepaid costs (such as the up-front costs of writing multiyear contracts). Such costs should be spread over the contract period, although, in some cases, they are spread over a fixed period (for example, three years in Italy).

There are, however, a number of issues with the calculation of reserves. One issue—in the case of outstanding claim reserves—is that they are actuarial estimates, and insurers can reduce their tax liabilities by increasing the reserves. In response, many countries require that such reserves be the amount calculated by an actuary and adopted by the insurer for financial reporting purposes. However, in general, large reserves are encouraged, if not actually required, by regulators. A second issue is that reserves for future payouts should be discounted so that they are not overstated and insurers undertaxed. However, regulators generally do not allow the discounting of reserves (since this decreases reserves), and most OECD countries do not require discounting for tax purposes. The exceptions are Australia, Canada, and the United States.

Some aspects of general insurance tax rules found in many countries are favorable to insurers because they allow deductions that exceed those that are required to properly measure annual income. Such deductions may be allowed as a general tax subsidy to the insurance industry or as tax assistance to help the industry cope with the volatility of its profits. The main forms of these are immediate deductions for prepaid expenses (Austria and Germany), equalization reserves for policies that have a volatile claim history (Austria, Germany, Luxembourg, the Netherlands, and Sweden), and catastrophe reserves for insured events that have a low probability of occurrence but that cause large financial losses if they do occur. Examples are insurance against earthquake and war risk (Belgium, Mexico, Switzerland, and the United Kingdom), and nuclear risk (Germany and the Netherlands).

International Tax Considerations

While international tax issues for general insurance are similar to those for life insurance, there is an even greater need to ensure that general insurance and other corporate profits cannot be moved offshore tax free by way of insurance contracts. If, as is likely, general insurance premiums are deductible business expenses, and if, also, premiums paid offshore are not taxable in the country in which the insured business operates, then the business profits of the insured can effectively be moved out of the country. Insurance can be a very flexible mechanism for shifting profits in this way. For example, almost anything can be the object of an insurance contract, arguably including risks with a high probability of occurring (declines in profit levels or asset values, for example), which can allow profits to be shifted offshore free of tax much like certain investment vehicles.

An alternative but similar way of moving profits out of the country is through captive insurance. Captive insurers are offshore insurance companies that are subsidiaries of domestic companies. Captive insurers were originally established to allow large companies to self-insure risks that commercial insurers would not assume (such as hazardous waste risks) while allowing the companies to receive a deduction for premiums when domestic tax rules disallowed a deduction for any reserves set aside to cover such risks. Aside from this commercial rationale, captive insurers have come to be used to transfer profits out of the domestic tax base, usually to a tax haven. Captive insurance can be used by both insurers and noninsurance companies in this way.

Two responses to the captive insurance issue are the U.S.-type rules that disallow a deduction for premiums paid to another company in the same group (that is, with common ownership or control) when there is no transfer of risk outside the group, and controlled foreign company rules that tax shareholders on their share of the income derived by a foreign company that is controlled or owned by a small number of domestic shareholders.

A complication for tax authorities in dealing with international insurance is the need to apply laws consistent with double-tax-agreement obligations. The OECD Model Tax Convention does not generally allow a source country to tax income derived by branches that do not operate in the source country through a permanent establishment. With insurance, substantial activity can take place without a permanent establishment. As a result, foreign insurers can structure their activities to avoid tax. Some countries attempt to deal with this problem by extending source taxation to insurance carried on through an agent. Australia and New Zealand exclude insurance from double-tax-agreement coverage.


Reinsurance plays a role in both general and life insurance but is especially important for the latter, where OECD data suggest it accounts for over 10 percent of gross premiums.6 Reinsurance is important to insurers because it enables them to manage and reduce their risks. This can be achieved by the insurer (or the direct writer) entering, in effect, into an insurance partnership with one or more reinsurers, thus sharing risks and income. Risk may also be managed by using nonproportional reinsurance arrangements that place a ceiling on the risks faced by the insurer. For example, under per-risk-excess-of-loss reinsurance, the insurer remains liable only for risks up to a maximum amount per claim. This enables an insurer to underwrite risks that it could not otherwise manage. Reinsurance can cover claims against the insurer, the timing of claims, investment risk, or expenses faced by the insurer.

In general terms, the tax rules for reinsurance should follow those that apply to the underlying insurance contract. That is because it is, in substance, an insurance contract where the policyholder is the direct writer. However, while reinsurance does play an essential commercial role in insurance risk management, tax rules need to take into account that reinsurance can be entered into largely for tax reasons. The most obvious example of tax-driven reinsurance is when it is used to transfer profits to a foreign reinsurer outside the domestic tax base. The international problems are even greater for reinsurance than for direct insurance because reinsurance contracts can easily be entered into without the presence of a permanent establishment or even a resident agent. They can thus easily fall outside the category of income taxable in the source country.

Reinsurance can also be used as a tax-efficient form of financing—financial reinsurance. For example, the direct writer reinsures policies on which the claims are reasonably predictable. The premium amount is, in effect, deposited with the reinsurer where it accrues investment income. When the claims are settled, the reinsurer pays the direct writer the premium plus the investment income. This functions much like any other investment except that for tax purposes, under normal rules, the premium or deposit is deductible, and the investment income accrues tax free.

Tax policymakers have responded to these issues in two ways. First, many countries levy withholding taxes on reinsurance premiums. Mexico does this on all reinsurance premiums. Australia, New Zealand, Spain, and the United States do so on premiums paid to nonresident reinsurers. Denmark achieves similar results by denying deductions for reinsurance premiums. It should be noted, however, that such taxes can be overridden by double taxation agreements, and they can act as a distorting tariff on legitimate reinsurance. Second, Australia and U.S. tax rules do not recognize, for deductibility purposes, reinsurance contracts that do not transfer insurance risk. An example would be where the reinsurer guarantees to return to the direct writer policy profits in excess of a set amount. These rules are, however, complex to administer, requiring a detailed examination of individual reinsurance agreements.

1OECD, Liberalisation of International Insurance Operations: Cross-Border Trade and Establishment of Foreign Branches (Paris, 1999).
2No double counting is involved, because the actuarial reserve is adjusted for actual claims.
3This makes transitional provisions especially important when contemplating a change to tax laws affecting the insurance industry. If insurance policies were written on the basis of one law, changing the law could undermine the profitability of those policies. It is generally desirable to design transitional rules that avoid windfall gains and losses, especially if the losses threaten the solvency of insurers.
4See David S. Holland and Denis Normand, Taxing Insurance Companies, OECD Tax Policy Studies No. 3 (Paris: OECD, 2001), pp. 77-78.
5The United States taxes policyholders on benefits received on maturity or surrender net of premiums and policyholder dividends. Death benefits are exempt, however.
6OECD, Insurance Statistics Yearbook, 1990-97 (Paris, 1999), Table VII.3.

    Other Resources Citing This Publication