Printer Friendly

Asymmetric information, captive insurers' formation, and managers' welfare gain.

Asymmetric Information, Captive Insurers' Formation, and Managers' Welfare Gain

Abstract

The generally accepted argument is that captive insurance firms are synergistic because they allow tax arbitrage and other cash flow increases which would positively impact the stocks of their parent corporations. This article presents asymmetric information and wealth transfer potential as additional arguments for the creation of insurance subsidiaries. It then studies the stocks of firms that owned captive insurers around the date of incorporation of their subsidiaries. The asymmetric information hypothesis which predicts that captive formation has no effect on the value of parent firms' stocks cannot be rejected.

Introduction

A captive insurer is an insurance subsidiary created to insure the accidental losses of its parent company. Bawcutt [1] reports that captive insurers have existed for several centuries. According to Fenske [12], self-insurance by companies dates back to the seventeenth century when British ship owners pooled their resources together to provide their own marine insurance. Captives did not become a major topic of interest in the insurance industry until the past three decades when the formation of insurance subsidiaries accelerated. Strazewski [26] reports that captives' share of the insurance market went from almost nil to 7 percent from 1973 to 1979. The increasing number of corporations creating their own captive raises the question, what economic benefit do captive insurers provide that is not available through commercial insurers?

In the fields of insurance and risk management, Bickelhaupt [2], Glenn, Cooper, and Rosenbaum [13], Nielson [17], Bouchard [4], and Doherty [9] give several reasons why firms choose to retain risk through a captive insurance program that can be grouped under two broad categories. The first one is synergy, which implies that firms with these programs experience an increase in cash flow. According to these authors, synergy takes place through factors such as cost reduction, tax advantages, and more efficient foreign coverage for firms with worldwide activities.(1) The second category is control. It deals with greater loss prevention that leads to lower loss experience. The effect of an increase in cash flow and decrease in loss experience would be to augment the value of corporations which form captive insurers. Such an increase in value would justify the existence of insurance subsidiaries. Cross, Davidson, and Thornton [6] report test results that weakly concur with these synergistic arguments. Similarly, Wood, Glascock, and Bigbee [27] conclude that the formation of captive insurers is tax based. The authors argue that the tax advantage accrues only to firms that formed a captive before the Internal Revenue Service's Revenue Ruling 77-316 which significantly reduced the tax benefits for some types of insurance subsidiaries in 1977.

The main contribution of this article follows from two arguments for the formation of captives that have been ignored in most of the insurance and risk management literature dealing with captive insurers. One of the arguments is inferred from the literature of asymmetric information and moral hazard. The other deals with the question of wealth transfer. According to asymmetric information, the formation of captives is an attempt by corporations to reduce the effective level of loss sharing thrust upon them in the direct insurance market. Wealth transfer maintains that a captive may be a tool to transfer wealth from one class of parent firms' claimholders to another.

This study first discusses the arguments for imperfect information as the origin of the formation of captive insurers. Second, it shows the implication of insurance subsidiaries on the value of the stocks of their parent corporations. Third, it examines empirically the effect of the creation of captives on the wealth of the stockholders of parent firms. This effect is measured through the application of the event-study methodology. The standard cumulative abnormal return (CAR) test based on the method of Fama, Fisher, Jensen, and Roll [11] is run to measure excess return to captive insurance firms; then the multi-variate regression model which is a variant of the seemingly unrelated regressions (SUR) of Zellner [28] applied to event studies by Binder [3] is used as a check for the CAR results. Finally, a conclusion is drawn about the economic justification for captive insurers.

(1)For detailed discussion of the tax arguments for captive formation refer to [13].

Arguments From Imperfect Information

Captive Insurers as the Result of Imperfect Information Under asymmetric information, it is prohibitively costly for commercial insurers to assess the true distribution of potential losses of their multitude of client firms, while this distribution is readily apparent to the managers of the insured corporations.

Rothschild and Stiglitz [22] model a competitive insurance market with asymmetric information where there exists one class of customers rated low risk and another class labeled high risk. These customers know their own accident probabilities which are unknown to the insurers. They demonstrate that in such a world price-quantity competition prevails. Contracts may simultaneously specify the price and quantity that can be bought rather than quoting price only, as a result of moral hazard. The quantity of insurance insureds can purchase is limited; thus they share some of the losses with the insurer. Another property of their model's competitive equilibrium is that low risk insureds suffer welfare loss specifically because of the presence of high risk customers in the market. The low risk class would be better off if there were no high risk insurance purchasers in the market, or if these risky individuals revealed their true loss distribution. The high risk class does not improve its plight whether it is alone in the market, lumped indiscriminately with low risks, or reveals its true riskiness.

Shavell [23] and Holmstrom [14] support the idea that under asymmetric information, non-Pareto optimal equilibrium leads to partial coverage against losses, explaining why an insurance policy with deductibles is optimal within this context. However, Shavell warns that the partial coverage against loss may subject the policyowner to some unwanted risk exposure.

Eisen [10] also supports the idea that departure from optimal risk sharing is a necessary incentive for getting the appropriate behavior from the insured. But an excess burden is put on the policyowner for the outcomes that are only partially its responsibility. Eisen states that outcomes depend on both the state of nature and the policyowner's actions; but the consequences are totally upon the policyowner because the insurer is unable to clearly assess the consumer's risk when asymmetric information holds. The result is that some policyowners assume more risk than they should.

Thus, in the context of asymmetric information there exists a clear incentive for a class of customers to single itself out of the crowd of insurance buyers in an attempt to reduce its level of loss sharing, or alternatively, as a means of getting more insurance for its money, leading to increased welfare. Captive insurance firms appear to be one mechanism for achieving that purpose. Functions of Captive Insurers

Captive as Information Providers at Minimal Costs: The captive insurer arrangement solves the problem of the cost of information. The probability distribution of losses of the parent firm is readily available to the captive insurer, and all intentions for altering it are presumably known, assuming the captive and its parent are under the supervision of the same management. Consequently, the premium assessed by the captive insurer is exactly a function of the true risk of the insured parent. When the captive transfers the risk through reinsurance, it provides the reinsurer with the probability distribution of losses of its parent at no, or at least non-prohibitive, information costs. This allows the parent corporation to purchase the adequate amount of insurance at a fair rate, up to the amount of losses it actually desires to retain. Thus, a captive may be viewed as the device that enables the insurance market to get the relevant information about the insurance-buying firm at a reasonable cost, and the reinsurance market provides the mechanism through which this information can be acted upon to serve signaling firms at a fairer rate than the direct commercial market's.

Captive Insurers and the Potential for False Signaling: It may be argued that if creating a captive is a signaling device that fosters the perception of low risk, high-risk firms can simply form an insurance subsidiary to take advantage of the favorable conditions offered in the reinsurance market. If high risk firms act accordingly, then the inability to distinguish among risk classes and the problem of prohibitive costs of information is merely transferred from the commercial market to the reinsurance market. However, this argument does not hold when the similarity between the decision to form a captive and the choice of debt level in the Ross [21] model of decision making under asymmetric information is established.

Within Ross's two-period model framework, firms which form captives are accountable at some future date for the decision they make currently. By creating an insurance subsidiary, a firm is perceived to belong to the low-risk class and it is assessed the premium corresponding to that class. However, if loss experience in the following period requires a disbursement larger than warranted by the probability of losses signaled originally by the parent company, the reinsurer charges a premium to match the true loss distribution plus a penalty for deception. That is, in the reinsurance market, the payment of premium at the end of the contract period provides a control mechanism that involves a built-in incentive for parent firms to signal the truth. Also, establishing a captive requires the parent to incur some costs well before the reinsurance process. These costs must be based on sound economic ground because of the potential for capital loss otherwise. Consequently, these initial set-up costs may dampen the incentive for false signaling and they may serve to establish the credibility of the signal.(2)

In summary, captive insurers signal low risk firms and may be an attempt to separate them from the riskier ones. They facilitate the supply of accurate information to the insurance market at minimal cost because they are a mere extension of their parent firms. This enables low-risk corporations to get the appropriate insurance coverage for the amount of premium paid. The costs required to set up a captive, coupled with the pattern of paying premiums at the end of the contract period in the reinsurance market, keeps the signal honest and bars high-risk firms from using captive insurers.

(2)Dionne [8] argues that in a competitive insurance market, one way to control the possibility that insured units move from insurer to insurer and declare false risk class, time after time, is to ask the insureds to make a high down payment of money when approaching an insurer for a first insurance contract. This amount of money is paid back at the end of the contract only if the insured told the truth. The set-up cost of a captive insurer can be viewed as a down payment, playing a role in mitigating false signaling in the reinsurance market.

Captive Formation and Stockholders' Wealth

Reagan and Stulz [19] demonstrate that even in the absence of agency problems, it is optimal for shareholders of value maximizing firms to issue labor contracts which induce workers to bear some risk unless all risk is diversifiable or the market price of risk-bearing services is zero. In that context, some conflict may arise between managers and shareholders because of the difference in the nature of risk exposure, according to Smith and Watts [25]. In general, managers have a large fraction of their wealth in the form of human capital tied to some specific firm. As a consequence, managers are more concerned with the total risk of firms' cash flow, unlike stockholders who focus only on systematic risk because they have access to the capital market and they can hold a well-diversified portfolio. given that managers are risk averse, it is rational for them to embrace measures that reduce the total variability of their firms' value. Hence, it is likely that whenever they perceive the direct insurance market to force on them a level of risk retention disproportionate with the amount that is justifiable by the expected probability distribution of losses of their firms, they attempt to correct the discrepancy, given the type of risk they bear.

So, one fundamental function of captive insurers appears to be to reduce the loss-sharing scheme implied by asymmetric information. They are instruments for reducing the risk exposure of their parent corporations. As such, they contribute to the reduction in the variability of parent firms' cash flow and reduce unsystematic risk. Stockholders can perform this service themselves at negligible cost by holding a well-diversified portfolio. Thus, stockholders do not value such a service. As a result, they are indifferent to whether a captive insurer is formed or not, and the value of their stocks remains unchanged around the date of formation of insurance subsidiaries.

The wealth transfer argument suggests that captives may be used to switch wealth from one class of their parents' claimholders to another. Kim, McConnell, and Greenwood [15] showed evidence that captive finance firms play such a role. Porat [18] suggests that the group of claimholders which may use captive insurers for personal gain are managers. Thus, captives may serve as tools for increasing the welfare of parent firms' managers at the expense of reduced stock value for corporations' residual owners.

The synergy argument implies that captives increase the value of parent firms and the value of stocks increases unambiguously as stockholders who are residual owners get the benefit of the increase in value.

These three views of the formation of captives are hypothesized as:

(1) Asymmetric information: the value of parent firms' stocks is invariant to the formation of captive insurers; (2) Wealth transfer: the value of parent firms' stocks decreases when captives are formed; and (3) Synergy: the value of parent firms' stocks increases when captives are formed.

Data and Methodology

The empirical investigation is conducted on two samples of firms listed on the New York Stock Exchange.(3) The domestic sample contains 21 firms owning a captive insurer domiciled in the U.S. The off-shore sample includes 70 parent companies with captives domiciled in Bermuda. The source of the samples is the Captive Insurance Company Directory 1983 issued by the Risk Planning Group, Inc. [20]. All parent companies selected are single-owner parents. The dates of incorporation of the off-shore captives were collected at the Insurance Division of the Department of the Registrar of Companies of Bermuda. The dates of incorporation for domestic captives were obtained by telephone survey and from the Best's Insurance Reports. Common stock data are obtained from the CRSP daily returns file and daily index file.

The effect of the creation of captive insurers on the stocks of their parents is studied initially. Then, additional tests that take into account Revenue Ruling 77-316 are conducted. Revenue Ruling 77-316 officially denies the tax deductibility of insurance premium payments made by single owner parents to their foreign captive insurers. Since the ruling may affect parent firms' net cash flows, the purpose of the tests is to verify whether, around the formation date, a difference in behavior exists between the stocks of corporations which formed captives before the ruling and the stocks of those which created their subsidiaries later. These additional tests are conducted on the off-shore sample only, since the ruling applies to foreign-based subsidiaries. Forty-six firms in the sample formed their captives before Ruling 77-316 was enacted on August 29, 1977, and 24 firms formed captives later. The Cumulative Abnormal Return Method

It is assumed that the ex-ante expected return for stocks is generated by the following return generating process commonly known as the market model: N/R sub jt = a subj + b subj 0 N/R sub mt + N/E sub jt Where R(jt) = rate of return on security j at time t, a(j), b(j) = constant parameters, R(mt) = rate of return on market portfolio at time t, N/E sub jt E(e(jt) = 0, Cov(R(jt),e(jt)= 0,Cov(e(it),e(jt)) = 0 for i j.

(3)An off-shore sample and a domestic sample were used separately because of arguments that legal constraints and supervisory authorities are more favorable to captives in foreign countries than in the U.S. This was an attempt to isolate those differences in the results.

The parameters a(j) and b(j) are estimated through the ordinary least squares technique over an estimation period extending from day -240 to +240 around the incorporation date of captives, but excluding the event period from day -45 to +45 around that date. The abnormal return over the event period is obtained as follows: e(jt) = R(jt) - (a(j) + b(j) R(mt)) Where R(mt) = return on the value weighted index at t, R(jt) = realized return on security j at t, e(jt) = residual for asset j at t.

The residuals are averaged across firms, for every period t to obtain the average residual (AR(t)) such that: AR sub t = (1/N) b summation of j from 1 to N b caret e sub jt

The CAR is computed by summing AR(t)'s over time. The CAR for a given day t within the event period is: CAR(t) = CAR(t-1) + AR(t)

The CAR over the event period is obtained as follows: CAR(1) = summation of t from f to 1 b AR sub t Where f stands for the first day of the event period, and 1 represents the last day.

Test Statistics: Following Brown and Warner [5] the test statistic for the null hypothesis of no abnormal return to stockholders during the event period is: CAR sub 1/b square root of 91 over 1/N {summation of j from 1 to N [1/388 (summation of t from -240 to -46 b AR' + summation of t from 46 to 240b AR')] raised to 1/2} Where AR' = (caret e sub jt - bar e) raised to 2 With bar e = 1/390 (summation of t from -240 to -46 b caret e sub jt + summation of t from 46 to 240 b caret e sub jt)

The statistic above is distributed t with 389 degrees of freedom.

The t-statistics for no abnormal return to stockholders during any event day t are obtained by replacing the numerator in (6) with AR(t).

The Multi-Variate Regression Model

The multi-variate regression model (MVRM) is a variant of the SUR technique applied to event studies by Binder [3]. The model consists of a system of return equations, one equation for each security j(j=1...n). The estimation is conducted with 481 daily stock returns for each firm in both the off-shore and domestic samples. These 481 returns include 240 daily returns before the date of formation, the return on the date of formation, and 240 daily returns following the date of incorporation. The day of captive formation is labelled t(0). The event period spans 91 days from t(0)-45 to t(0)+45.

The event period in the MVRM analysis is broken down into three sub-periods to correspond with relevant ranges of the CAR results.(4) The three sub-periods are defined as the pre-formation, the formation, and the post-formation sub-periods. If s stands for sub-period then: for the pre-formation sub-period, s = 1 and it covers t(0)-45 to t(0)-24, for the formation sub-period, s = 2 and it spans t(0)-23 to t(0)+5, for the post-formation period, s = 3 and it covers t(0)+6 to t(0)+45.

The system of equations allows the excess returns to differ in sign and size across firms and across event sub-periods, and it can be expressed as follows: R(1t) = a(10)+a(11)D(0t)+b(10)R(mt)+b(11)R(mt)D(0t)+ D(st)+e(1t) R(2t) = a(20)+a(21)D(0t)+b(20)R(mt)+b(21)R(mt)D(0t)+ D(st)+e(2t) . . . R(nt) = a(n0)+a(n1)D(0t)+b(n0)R(mt)+b(n1)R(mt)D(0t)+ D(st)+e(nt) Where R(jt) = return on security j at time t(j=1...n), D(0t) = dummy variable that accounts for change in intercept and

slope at the beginning of the event period (day t(0)-45) and

beyond, R(mt) = return on value-weighted market index, D(st) = dummy variable that accounts for excess return during the

event sub-period s,

= excess return to firm j during event sub-period s, and e(jt) = disturbance term, independently and identically distributed

within each equation, but contemporaneous covariance across

equations may be non zero.

(4)Relevant ranges are functions of the range over which the CAR test found significant abnormal return.

Tests: Hypotheses about excess returns were tested using linear constraints on the coefficients of the dummy variables D(st).

The following tests are run:

Test 1: = 0; the excess return during all event sub-periods equals zero

for firm j.

Test 2: = 0; the cumulative excess return over the formation

sub-period equals zero.

Test 3: = 0; the cumulative excess return over the pre-formation

sub-period equals zero.

Test 4: = 0; the cumulative excess return over the post-formation

sub-period equals zero.

Test 5: =0; the cumulative excess return equals zero over the 91 days

event period for firm j.

Results CAR Results

Off-Shore Captives: The effect of the creation of off-shore captives on the stocks of parent companies is reported in Table 1. The AR return on day zero is -.0025, with a t-value of -1.2151 which is not significant at the 5 percent level. However, the AR return two days prior to incorporation is significantly negative, though.

The CAR return over the whole event period is -.026259. The t-value is -1.336. The wealth of stockholders is not affected 90 days around the date of creation of an off-shore captive.

The CAR for the first 46 days of the event amounts to -.014543 with a t-value of -1.0407. The CAR is mostly on the negative side and it appears fairly random. However, a closer look at the CAR reveals a continuous negative trend which starts from day -40 to about day -23.(5) The CAR that accrues to stockholders over these 17 days amounts to -.018603, with a significant t-value of -2.1898. The CAR to stockholders over the 23 days from t(0)-23 to t(0) is .003251 and not significant.

(5)Attention was drawn to this downward sloping segment of the CAR because of some prior information from Appleby, Spurling, and Kempe a Bermudian law firm involved in the incorporation process of captive insurers. The head of the captive department revealed that a public announcement of the intention to incorporate captives is required in the Royal Gazette, an official newspaper of Bermuda. She added that the average length of time between the announcement and the date of incorporation is approximately four to six weeks. If so, there exists potential for leakage of information about 20 to 30 days before the actual incorporation date in the case of Bermudian captives. This downward sloping segment could have been a manifestation of such a leakage.

Domestic Captives: Table 2 reports the effect of the creation of domestic captive insurers on the stocks of their parents. The AR return on the date of incorporation of the captives equals -.0062106 with t-value of -1.60000. Stockholders gain no abnormal return on the date of incorporation of domestic captives.

The CAR return over the 91-day period is -.015081 and t-value equals -.40728. As with the off-shore sample, stockholders do not earn abnormal returns for the whole event period. However, the CAR drops steadily from .018030 to -.026382 between day t(0)-19 to t(0)+2; afterward, the CAR remains close to -.020 on average for the remainder of the event period. The negative CAR earned by stockholders during the 22 days between event day t(0)-19 to t(0)+2 equals -.044412 with a t-value of -2.43935, which is significant. The CAR from event day t(0)+1 to t(0)+45 amounts to -.019363 with a non significant t-value of -.7355. Stockholders appear to earn significantly negative abnormal return during 20 days preceding the incorporation of domestic captives. The MVRM Results

Off-shore Captives: The results of Test 1 reported in Table 3 show that 12 firms have significant excess returns in some event sub-period. For the majority, there is no significant impact on stocks, in any of the sub-periods. The tests reveal that during the pre-formation sub-period, six out of nine significant excess returns are negative. Consequently, the abnormal return found by the CAR method over this same sub-period may be due to these six firms' pulling the cumulative abnormal return on the negative side for the whole sample. Test 3 reported in Table 4, investigates that matter, and it checks whether the cumulative excess return is significantly different from zero over the pre-formation sub-period. The MVRM technique finds a negative cumulative excess return like that of the CAR method. But unlike the CAR which finds the negative abnormal return significant at 5 percent, the MVRM's estimates are not different from zero at the level of significance.

Test 2 and 4 in Table 4 show that the cumulative abnormal return over the formation and the post-formation sub-periods are not significant. These results are consistent with the CAR findings.

Test 5 in Table 5 reveals that only five firms show a significant CAR over the 91 days of the event period. Also the joint test that all the firms together earn zero cumulative excess return over the 91-day event period cannot be rejected.

Domestic Captives: The MVRM tests conducted on the domestic sample show that five out of 21 firms earn significant excess returns over some event sub-period. They also find a negative but not significant cumulative excess return over the sub-period for which the CAR found a significant negative return. They reveal no cumulative abnormal return for either the preformation sub-period or the post-formation one. Furthermore, the MVRM cannot reject the joint test that all firms together earn zero cumulative excess return over the 91-day even period. The off-shore sample, like the domestic one, exhibits no change in the value of the stocks of parent firms.(6)

In summary, the MVRM technique shows that the overwhelming majority of firms do not earn any significant excess return. The MVRM tests cannot reject the hypothesis that in aggregate, the creation of captive insurers by the samples of corporations studied has no impact on the stocks of parents around the date of incorporation of the captives. Off-shore Captives and Revenue Ruling 77-316 Effects

Tables 6 and 7, respectively, report the results of the CAR applied to the 46 firms that formed their captives before Ruling 77-316 and the 24 corporations which created theirs after August 1977. Firms that formed their insurance subsidiaries before Revenue Ruling 77-316 earn no significant excess return on the date of incorporation of their captives. The CAR over the 91-day event period is -.034356 with t-value of -1.4449. For this group of firms, there is no evidence of a significant abnormal return related to the formation of captives.

Corporations that formed their insurers after August 29, 1977 exhibit a significant excess return of -.00829 (t=-2.275) on the date of incorporation. Furthermore, this group earns a significant negative return two days before the date of formation, while it obtains a significant positive return six days before and three days after. The shifting signs of these returns do not allow any clear-cut interpretation of the negative observation on the day of formation. The CAR for this group is -.01074 (t=-.30886) over the 91-day event period and its plot is fairly random, i.e., it does not drop on the day of incorporation and remain at that lower level thereafter.(7)

Whether the date of formation is before or after ruling 77-316, the tests cannot reject the hypothesis that stocks of companies that create captives do not earn abnormal return around the incorporation date of the subsidiaries.(8) Managers as Potential Beneficiary of Increased Welfare

Stockholders' wealth does not seem affected by captive insurers formation. However, captives are still justified if they improve the second best equilibrium, and they move the competitive insurance market one step closer to the Pareto optimal equilibrium by promoting truthful disclosure. Also, (6)Tables containing the findings of the MVRM on the domestic sample are not reported since the tests do not lead to conclusions different from the ones reached with the off-shore sample. Nevertheless, they can be made available on request. (7)Plots for Tables 6 and 7 can be made available on request. (8)A study by Cross, Davidson, and Thornton [7] concludes otherwise. To splittheir sample, their study uses December 26, 1978, the date when the tax court ruled against Carnation, as opposed to August 29, 1977 when the IRS passed the ruling against the deductibility of premiums paid to single-owner captives. captive insurers possess the attributes for enabling less diversified claimholders of corporations to extract benefits accruing from lower cash flow variability that results from greater insurance coverage.(9) That is, they ma serve as tools for reducing agency costs to the less diversified claimholders of corporations. Smith and Warner [24], and Mayers and Smith [16] explain the purchase of insurance by firms as an activity for reducing agency costs and that involves stockholders contracting relationships with bondholders, employees, customers, suppliers and managers. Beyond that, captive insurers can provide managers with opportunities for greater appropriation of their (9)A captive provides opportunity for expanded responsibility in the form of the supervision of an entirely new division. The incremental responsibility will command pecuniary reward, possibly greater clout within the firm for the managers in charge of the insurance subsidiary. Perquisites coming from higher prestige can be picked, too. corporations' cash flows, which may make these subsidiaries a favorite method for acquiring the additional coverage implied by truthful disclosure.(10)

Conclusion

It is generally argued that captive insurance firms are synergistic because they allow tax arbitrage and other cash flow increases. Synergy implies greater wealth for the stockholders of firms that form captives. Besides synergistic factors, this article presents asymmetric information and wealth transfer

(10)Low-risk firms could choose a loss-retro scheme as a means of acquiring proper coverage. However, loss-retro lacks the key feature provided by captives. It does not offer managers an opportunity for appropriation of wealth. Note also that loss-retro does not require the up-front costs needed to set captives; thus it does not have the signaling attribute of captives, and it may be used more freely by high risk firms. As such, loss-retro does not contribute to improving the second-best equilibrium in a competitive insurance market, while captives may.

Table : Off-shore Captives: Stock CAR Around Date of Formation

Table : Domestic Captives: Stock CAR Around Date of Formation

Table : Off-shore Sample, Stock: MVRM Test 1 H(0): Excess Return During All Event Sub-Periods Equal Zero for Firm j

Table : Off-shore SAmple, Stock: MVRM Test 2, 3, 4 H(0): Cumulative Excess Return Over Each Event Sub-Period Equals Zero

Table : Off-shore Sample, Stock: MVRM Test 5 H(0): Cumulative Excess Return Over 91-Day Event Period is Zero for Firm j

Table : Stock CAR Around Date of Formation, Before I.R.S. Ruling 77-316

Table : Stock CAR Around Date of Formation, After I.R.S. Ruling 77-316 potential as arguments for the creation of insurance subsidiaries. It supports that the formation of captive insurers may evolve from the existence of imperfect information in the direct insurance market. Captives provide the vehicle for revealing true information at minimal cost, thus enabling low-risk firms to signal their true loss distribution. By signaling, low-risk corporations can reduce the level of loss sharing imposed on them in the direct insurance market. The reinsurance market provides the arena for readjustment of coverage and its premium disbursement scheme, coupled with the up-front set-up cost of captives, keeps the signal honest. Cumulative abnormal return and multi-variate regression model techniques were used to analyze stocks of parent firms around the date of incorporation of their insurance subsidiaries. Overall, the value of stocks of the parent corporations remains unchanged. Samples in this study provide no evidence of synergistic gain, nor wealth transfer.(11) The asymmetric information hypothesis that the value of parent firms' stocks is invariant around the formation date of their captive insurers cannot be rejected.

From the empirical results in this study, the contention is that, as expected, the primary function of captive insurers is to provide insurance, especially, larger amounts of coverage, to their parents. The welfare gain derived from the creation of captives most likely goes to the managers of parent firms. Finally, this research may be extended by increasing the sample size, and pinpointing the dates when the formation of various captives were first publicized in an attempt to improve the power of the tests and possibly gain additional insights. An avenue for further empirical exploration of the subject would be a direct investigation of managers in charge of risk management, before and after the creation of captives and/or the difference between risk managers of firms with captives versus those of firms without.

(11)Plots of the CAR drift mostly on the negative side, not significantly, though. Stockholders seem to sustain a decrease in wealth, albeit negligible. Note that an amount negligible for all stockholders together may be significant for a handful of managers.
COPYRIGHT 1989 American Risk and Insurance Association, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Diallo, Alahassane; Kim, Sangphill
Publication:Journal of Risk and Insurance
Date:Jun 1, 1989
Words:5586
Previous Article:A historical perspective on the present value assessment of medical care.
Next Article:Compensation alternatives for occupational disease and disability.
Topics:

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters