reinsurance section, issue no. 52, august 2003 reinsurance news · 2012-01-19 · 2 • reinsurance...

REINSURANCE NEWS

ISSUE NO. 53 • MARCH 2004

Embedded Derivativesand Financial Reinsuranceby Larry Carson

Implications of aConsolidating Marketplace A report from an ACLI AnnualConference Sessionby Hank Ramsey

NEWSLETTER OF THE REINSURANCE SECTION

continued on page 4 continued on page 11

Disclaimer

The following paper sets out the approach beingused by Reinsurance Group of America (RGA) toapply SFAS 133 Implementation Issue B36 to

coinsurance funds-withheld and modified coinsurancetransactions that are classified as “financial reinsur-ance.” The information in this paper is provided only forinformation purposes and is not intended and shouldnot be construed as accounting, auditing, legal or taxadvice with respect to any specific facts or circum-stances, as the facts and circumstances at othercompanies may be different materially from those atRGA and may result in different conclusions. RGAmakes no representation or warranty as to the accuracyor completeness of the information provided herein, andyou may not rely for any purpose on any ideas, judg-ments, opinions or analyses provided in this paper. Youare encouraged to consult with your accountants, audi-tors, legal and other professional advisors to determinethe proper course of action for your company in connec-tion with the matters discussed in this paper.

Abstract

Financial reinsurance transactions contain twoembedded derivatives as defined under B36: one withinthe funds-withheld asset and the other within the expe-rience refund provision. The net of these two embeddedderivatives, which is what must be placed at marketvalue on the GAAP balance sheet, is zero at all points intime at which the transaction continues to be consid-ered financial reinsurance.

Background

This white paper sets out a proposed application ofSFAS 133 Implementation Issue B36 to coinsurancefunds-withheld and modified coinsurance transactionsthat are classified as “financial reinsurance.”

B36 requires the identification, bifurcation and valu-

What will be left of the reinsurance marketplacewhen the consolidations are over? That ques-tion and others were addressed by a session

featuring an S&P analyst, a pricing actuary and a rein-surance executive on October 14, 2003 in Miami whenthe American Council of Life Insurers held their annualconference. The consensus was that direct writers havebecome “hooked on reinsurance” in recent years, and arefeeling some pain as reinsurers consolidate. The remain-ing reinsurers are not bidding as aggressively,particularly for business that is not as profitable as theywould like it to be.

Rodney Clark, a director at S&P, led off with hisassessment of the market. He showed how the markethas become much more concentrated in the last sixyears. In 1997, 16 reinsurers wrote 90 percent of themarket. Today, that number is down to 11 reinsurers,and he estimated that we may be down to six to eightreinsurers by the end of 2005. Mergers and acquisitionsaccount for most of the decline in reinsurers. Mr. Clarkrecited a quick list of transactions, based on 1997 rankings:

• #1 ERC bought #8 Phoenix Re, and then #11 AULRe

• #3 RGA Re bought #10 Allianz Re• #5 Swiss Re bought #6 Life Re, and then #7 Lincoln

Re• #9 Guardian has put their reinsurance business in

runoff• #15 Munich Re bought #16 CNA Re[Subsequent to the conference, ERC announced that

it was selling the old Phoenix Re business and placingtheir remaining life reinsurance operations in run-off.]

Mr. Clark said there are many reasons for the consol-idation. Some companies have exited reinsurance as aline of business; others have succumbed to financialdistress, capital strain or lack of scale. With the attrac-tive margins available in the current hard P&Creinsurance market, access to capital has been limitedfor life reinsurers that are part of multi-line reinsur-

2 • RE INSURANCE NEWS • March 2004

REINSURANCE NEWSNEWSLETTER OF THE REINSURANCE SECTION

NUMBER 53 • MARCH 2004

This newsletter is free to section members. A subscription is $15.00 fornonmembers. Current-year issues are available from the PublicationsOrders Department. Back issues of section newsletters have beenplaced in the Society l ibrary, and are on the SOA Web site,www.soa.org. Photocopies of back issues may be requested for anominal fee.

Expressions of opinion stated herein are, unless expressly stated to thecontrary, not the opinion or position of the Society of Actuaries, itssections, committees or the employers of the authors.

The Society assumes no responsibility for statements made or opin-ions expressed in the articles, criticisms and discussions contained inthis publication.

Newsletter EditorDean S. Abbott, FSAING Re • Group Life, Accident and Health Reinsurance20 Washington Avenue SouthMinneapolis, MN 55401Phone: (612) 342-3815 • Fax: (612) 342-3209E-mail: [email protected]

OFFICERSChairpersonMelville J. Young, FSA, MAAAVice-ChairpersonMichael E. Gabon, FSASecretary/TreasurerTimothy J. Tongson, FSA

Council MembersDean S. Abbott, FSAJay D. Biehl, FSARonald L Klein, FSARichard K.M. Lau, ASAJohn O. Nigh, FSAHenry B Ramsey, III, FSALarry Warren, FSA

SOA StaffLois Chinnock, Section ManagerPhone: (847) 706-3524 • Fax: (847) 273-8524E-mail: [email protected]

Glenn Pinkus, DTP CoordinatorPhone: (847) 706-3548 • Fax: (847) 706-3599E-mail: [email protected]

Society of Actuaries475 N. Martingale Road, Suite 600 • Schaumburg, IL 60173Phone: (847) 706-3500 • Fax: (847) 706-3599 • Web: www.soa.org

Copyright © 2004

Society of Actuaries • All rights reserved • Printed in the United States

of America

COVER STORY1 Embedded Derivatives and

Financial Reinsurance

Larry Carson

Implications of a ConsolidatingMarketplace: A report from anACLI Annual Conference Session

Hank Ramsey

3 Chairperson’s Corner

Melville J. Young

OPINION

12

19

20

23

30

International Financial Reporting

Standards and Insurance

Sam Gutterman

SAFE Pool Provides NewCatastrophic CoverageAlternative for Life InsuranceIndustry

R. Dale Hall

Capacity in the US Life InsuranceMarket- A View from the Top ofthe Pyramid

Michael DeKoning

The Recapture Provision

Larry Warren

Spring Meeting ReinsuranceSessions

John Nigh

FEATURES

March 2004 • RE INSURANCE NEWS • 3

One of the founding principles of theReinsurance Section was to supportthe educational process for actuar-

ies in the area. It was recognized thatin order to facilitate this process ourliterature needed to be periodicallyreviewed and enhanced.

The Re insurance Sec t i on wasfounded in 1982. The Section’s workinformally began approximately twoyears earlier when the SOA formed atask force. The charge of the task forcewas t o c ons t ruc t a s e r i e s o f wh i t epapers on various reinsurance topics.Until that time there was John Wooddy’swhite paper, which contained a generaldescription of life reinsurance and littleelse.

During the ensuing two years a numberof additional white papers were authoredby the task f o r ce . These in c luded Bill Tyler’s paper on reinsurance under-wr i t ing ; Tom Heaphy ’s paper on thereinsurance treaty; Lucie Cossette’s andDenis LaPierre ’s reconstruct ion o f theWooddy article; MikeWinn’s paper on rein-surance pricing and Court Smith’s paperon re insurance accounting. There wereseveral other papers and the taskforce alsoincluded our resident Harvard grammar-ian , Den is Lor ing, who amongs t o therthings policed our use of the split infini-tive. Since the formation of the section, thelife reinsurance world has been enrichedby the written work of a number of individ-uals. Most notably the book written byDenise Fagerberg and John Tiller entitledLife Reinsurance and the book written byDave Atkinson and Jim Dallas entit ledPricing Life Reinsurance.

Approx imate ly a year ago theReinsurance Section Council decided thatthis would be a good time for a fresh look

at reinsurance literature since more thantwo decades had passed since the last over-all effort. Mike Gabon has ‘volunteered’ toform a committee whose responsibility willbe to assemble, review and then fill in thegaps. During the past year several of youhave expressed an interest in participatingin this important project. If you have aninterest in participating please contactMike at [email protected].

P.S. As part of this effort we are tryingto locate the white papers that were writ-ten by the earlier task force. My copiesseem to be AWOL. If anyone has a copy Iwould appreciate hearing from you.??

Chairperson’s Cornerby Melville J. Young

Mel Young, FSA,

MAAA, is executive

vice president and

vice chairman for

RGA Reinsurance

Company located in

Norwalk, Conn. He

may be reached at

myoung @rgare.com.

continued on page 5


ation in all coinsurance funds-withheld andmodified coinsurance transactions of embeddedderivatives contained within those agreements.While there is considerable disagreement as towhat risks these embedded derivatives encom-pass, what is the “host contract” (in the parlanceof SFAS 133) and other related issues, there islittle doubt that B36 applies equally to allfunds-withheld and modified coinsurance trans-actions, whether or not they are classified as“financial reinsurance.”

Having said that, the application of B36 tofinancial reinsurance transactions is problem-atic at best. Taken literally, B36 requires theidentification, bifurcation and valuation of anembedded derivative within the funds-withheldor modco asset, which, for a financial reinsur-ance transaction, does not even appear on theGAAP balance sheet! A blind application of B36would not take into account the specializednature of financial reinsurance transactions,which current GAAP accounting recognizes ashaving little to no economic impact outside ofthe reinsurance fees collected. It is doubtfulwhether this would create greater transparencyon a company’s GAAP financials or lead togreater understanding of the economic results ofsuch transactions.

As we will argue below, however, a properapplication of B36 to financial reinsurancetransactions results in no net balance sheet orincome statement impact, as there are twocompletely offsetting embedded derivatives tobe found in such transactions.

Note that while the following analysis is froma reinsurer’s perspective, we believe that aceding company’s perspective should follow asimilar logic.

Introduction

For purposes of this discussion, “financialreinsurance” transactions are defined as rein-surance transactions (and related transactions)

that fail to meet the SFAS 113 test to beaccounted for as reinsurance under GAAPaccounting1. In essence, financial reinsurancetransactions are those where the likelihood ofrealizing a material, long-term economic loss islow.

We distinguish between two types of financialreinsurance transactions:

1. A “non-cash” financial reinsurance transac-tion is one in which the net cash flow to theceding company either at treaty inception orupon new business being added is equal to zero.Under current GAAP accounting, there are noassets or liabilities on the GAAP balance sheet,and the GAAP income consists of the feesearned under the reinsurance transaction. If netcash other than the fees changes hands undersuch a transaction—typically, this wouldhappen if a loss develops on the underlyingreinsurance, such that a loss carryforward isestablished (a fairly rare occurrence)—thensuch cash is accounted for the same way as cashceding commissions under a “cash” financialreinsurance transaction.

2. A “cash” financial reinsurance transactionis one in which the net cash flow to the cedingcompany either at treaty inception or upon newbusiness being added is positive. The GAAPbalance sheet shows an asset equal to the cashoutstanding—and nothing else—while GAAPincome consists of the interest and fees earnedon the cash outstanding.

Non-Cash Financial ReinsuranceTransactions

We will assume that we are working with ageneric coinsurance funds-withheld financialreinsurance transaction (the treatment for acombination coinsurance-modified coinsurancetransaction would be substantially the same).

1 This determination is made on an ongoing basis. If a financial reinsurance transaction experiences a significant change to its

risk profile, then this discussion may no longer apply, i.e., SFAS 133 DIG B36 might need to be applied to such a transaction.

Embedded Derivatives • from cover


continued on page 6

Embedded Derivatives…

For an accounting period t, we define thefollowing terms:

PRGt = P – B – E – ∆V + rt • FWt-1

where PRG = preliminary reinsurance gainP = premiumsB = benefitsE = commission and expense

allowancesV = statutory reservert = funds-withheld interest rate

applicable to period tFW = funds-withheld balance.

Next, we define an experience refund foraccounting period t as follows:

ERt = max { 0, PRGt – Ft – LCFt-1 • (1 + it) }

where F = reinsurance feesLCF = loss carryforwardit = loss carryforward interest rate

applicable at time t= 3-month LIBORt + j

j = spread over LIBOR

with the further provision that ER may be set to0 at the option of the reinsurer after a certainpoint in time and/or upon other specified condi-tions2.

We also define the loss carryforward at time t as

LCFt = max { 0, LCFt-1 • (1 + it) + Ft – PRGt } .

Then, at any given point in time, the reinsur-ance cash settlement is defined as:

CS = PRG – ER .

At the point in time that experience refundsare set to 0 by the reinsurer, the cedingcompany is allowed to recapture the treaty byrepaying any current loss carryforward.

Application of B36

We are assuming, for purposes of this discus-sion, that this reinsurance transaction has beendetermined to be financial reinsurance, i.e.,there is a low probability of realizing a material,long-term economic loss. Put another way, sensi-tivity testing has indicated that, with highlikelihood, the ceding company will recapturethis transaction at the appropriate time. Sincerecapture entails the repayment of any losscarryforward, which, by its very definition accu-mulates any fees that were previously notcollected out of statutory profits, it follows thatnon-cash financial reinsurance transactions arethose with a high degree of likelihood that thereinsurer will collect the reinsurance fees andnothing more. 3

In other words, with a high degree of likeli-hood, the present value at treaty inception ofthe cash settlements (discounted at the short-term series of interest rates it) will be equal tothe present value of the reinsurance fees, i.e.,PV (CS) = PV (F). More broadly, under theassumption that we are testing the financialreinsurance transaction on an ongoing basis toensure that it still qualifies as financial reinsur-ance, we may say that, at any given point intime t,

PVt (CS) = PVt (F) + LCFt .

2 Actually, what we are calling the experience refund may be thought of as consisting of two pieces: a decrease in the relief

balance and a true refund of “excess” profits. In other words, what we are calculating above as ER is the amount of profits

available, and these may either (a) be used to amortize the relief (by increasing the funds-withheld balance); or (b) be

returned to the ceding company. In either case, the impact on the current accounting period’s cash settlement is the same,

since both the increase in the funds-withheld balance and an experience refund are items that the reinsurer pays in cash to

the ceding company. However, the division of this amount into these two components does impact the reinsurance settle-

ment items in future accounting periods, as it determines the beginning-of-period funds-withheld balance for the next

accounting period.3 Of course, while experience refunds are being paid, the reinsurer cannot collect any more income than the reinsurance fees.


Embedded Derivatives.. • from page 5

Substituting terms, we see that, with a high degree of likelihood,

PVt (F) + LCFt = PVt (CS)= PVt (PRG – ER)= PVt (PRG) – PVt (ER)

or, rearranging terms,

PVt (ER) = PVt (PRG) - PVt (F) - LCFt .

To be more specific,

(1) , where

dk=Π(1+im)-1=Π(1+LIBORm+j)-1

and w = time of recapture.4

Now, for any period t,

PRGt =P–B–E–∆V+rt•FWt-1

=P–B–E–∆V+ht•FWt-1+(rt - ht)•FWt-1

=Lt+Ht+EDt

where ht = interest rate for period t on host contractLt = liability cash flows for period t

= P–B–E–∆VHt = host contract interest for period t

= ht•FWt-1

EDt = embedded derivative cash flows for period t= (rt - ht)•FWt-1

Substituting into equation (1) above, we get equation (2):

PV ERd

dER

d

dPRG

d

dF LCFt

k

tk

k t

wk

tk t

w

kk

tk t

w

k t( ) = = − −•

= +

•

= +

•

= +∑ ∑ ∑

1 1 1

k

m=1

k

m=1

4 One may question what is the proper set of discount rates to be used in computing present values. As the

following analysis will show, using the discount rates LIBOR + j, where j is the spread over LIBOR used in

calculating the treaty loss carryforward, leads to a value of 0 for the embedded derivative at treaty inception

(indeed, at all times for as long as the treaty is still sufficiently profitable to be considered financial reinsur-

ance). This is because, under most scenarios, the present value of future cash flows associated with the

embedded derivative, discounting at LIBOR + j, will be 0. Since the embedded derivative needs to have a

value of 0 at treaty inception, this implies that discounting at LIBOR + j is correct.



continued on page 8

(2)

Next, we note that

ERt = max {0,PRGt–Ft–LCFt-1•(1+it)}= max {0,Lt+Ht+EDt–Ft–LCFt-1•(1+it)}.

Given that any existing loss carryforward is paid at the time of recapture5, we know that LCFw = 0.

Thus,

Substituting into equation (2) above, we get equation (3), for any point in time t:

(3)

Finally, when considering the present value of future cash settlements at any point in time t, wearrive at equation (4):

(4)

5 Note that this analysis does not require the ceding company to recapture the reinsurance agreement at the

time at which it is most advantageous to do so. It merely assumes that, with very high likelihood, the ceding

company will recapture at some point in the future. In other words, even if the ceding company does not

recapture the transaction at the point in time when economic analysis would suggest that it is in its best

interest to do so, sensitivity testing would still show that, with very high likelihood, we expect them to do so

in the future.

PV ERd

dL H ED

d

dF LCF

d

dL H ED F LCF

tk

tk t

w

k k kk

tk t

w

k t

k

tk t

w

k k k k t

( ) ( )

( )

= + + − −

= + + − −

•

= + = +

•

•

= +

∑ ∑

∑1 1

1

LCF LCFd

dLCF

d

d

LCFd

dLCF

d

d

d

dLCF i LCF

t tt

tw

w

t

k

k t

wk

tk

k

t

k

tk

k t

w

k k

= −

= −

= + −

+ − •+ −

•

−

= +

•−

•

• −

= +

•

∑

∑

( )( )

( )

( ( ) )

1 11 1

1

1

1

1

1

1

PV ERd

dL H ED F LCF

d

dL H ED F

d

dLCF i LCF

d

dL H ED F LCF

tk

tk t

w

k k k k t

k

tk t

w

k k k kk

tk t

w

k k k

k

tk t

w

k k k k

( ) ( )

( ) ( ( ) )

(

= + + − −

= + + − − + −

= + + − −

•

= +

•

= +

•

= +

− •

•

= +

∑

∑ ∑

∑

1

1 1

1

1

1

kk k ki LCF− • + +1 1( ) )

PV CS PV PRG ER

PV PRG PV ER

d

dL H ED

d

dL H ED F LCF i LCF

d

dF LCF LCF

t t

t t

k

tk k k

k

tk t

w

k t

w

k k k k k k k

k

tk t

w

k k k

( ) ( )

( ) ( )

( ) ( ( ) )

( ( (

= −= −

= + + − + + − − + +

= − −

•

= += +

• − •

= +

• − •

∑∑

∑11

1

1

1

1

11 + ik)))


Embedded Derivatives... • from page 7

In other words, with a high degree of likeli-hood, the only items that matter in terms of thecash settlements are the reinsurance fees, aswell as the change in any loss carryforwardbalance, with interest.

What we then see, is:

1. There is an embedded derivative in boththe liabilities (i.e., within the preliminary rein-surance gain) as well as in the experiencerefunds.

2. On a present value basis, these embeddedderivative cash flows precisely cancel each otherout, with a high degree of likelihood.

We therefore conclude that, taking intoaccount both embedded derivatives, the marketvalue of the single, net embedded derivative is 0at any given point in time

6.

Cash Financial Reinsurance Transactions

For the purpose of this discussion, we willassume that we are working with a coinsurancefunds-withheld financial reinsurance transac-tion (the treatment for a combinationcoinsurance-modified coinsurance transactionwould be substantially the same).

For an accounting period t, we define the follow-ing terms:

PRGt = P–B–E–∆V+rt•FWt-1

where PRG = preliminary reinsurancegain

P = premiumsB = benefitsE = expense allowancesV = statutory reservert = funds-withheld interest

rate applicable to period t

FW = funds-withheld balance.

Next, we define an experience refund foraccounting period t as follows:

ERt = max {0, PRGt–EAt-1•(1+it)}

where EA = cash experience accountit = cash experience account

interest rate applicableat time t

= 3-month LIBORt + j

j = spread over LIBOR(which includes both aninterest component anda fee component)

with the further proviso that ER may be set to 0at the option of the reinsurer after a certainpoint in time and/or upon other specified condi-tions7.

We also define the cash experience account attime t as

EAt = max { 0, EAt-1 • (1 + it)– PRGt } .

Then, at any given point in time, the reinsur-ance cash settlement is defined as:

CS = PRG – ER .

At the point in time that experience refundsare set to 0 by the reinsurer, the cedingcompany is allowed to recapture the treaty byrepaying any unamortized cash experienceaccount.

Application of B36

We are assuming, for purposes of this discus-sion, that this reinsurance transaction has beendetermined to be financial reinsurance, i.e.,there is a low probability of realizing a material,long-term economic loss. Put another way, sensi-tivity testing has indicated that, with high

6Note that, per SFAS 133 Implementation Issue B15, there can be only one embedded derivative per hybrid instrument. In

other words, a reporting entity is required to net these two embedded derivatives against each other.7Here, since cash is changing hands—which is being kept track of via the experience account—an experience refund would not

be payable until the experience account had reached 0. This is what is commonly referred to as “full amortization,” since all of

the reinsurance gains are being used to amortize the experience account. Some transactions instead feature “scheduled amor-

tization,” where, assuming specified conditions are met, the amount of amortization of the experience account each

accounting period is limited by some pre-defined formula, and any profits in excess of those being used to amortize the expe-

rience account are returned to the ceding company as an experience refund.


likelihood, the ceding company will recapture this transaction at the appropriate time. Since recap-ture entails the repayment of any cash experience account, which, by its very definition, accumulatesany cash and fees on that cash relief that were previously not collected out of statutory profits, itfollows that cash financial reinsurance transactions are those with a high degree of likelihood thatthe reinsurer will collect its cash investment, interest and reinsurance fees on that cash, and noth-ing more.

In other words, with a high degree of likelihood, the present value at treaty inception of the cashsettlements (discounted at the short-term series of interest rates it) will be equal to the initial cashceding commission. More broadly, under the assumption that we are testing the financial reinsur-ance transaction on an ongoing basis to ensure that it still qualifies as financial reinsurance, we maysay that, at any given point in time t,

PVt (CS) = EAt .

In other words, at time t,

EAt=PVt(CS)=PVt(PRG–ER)=PVt(PRG)-PVt(ER)

and thus

PVt (ER) = PVt (PRG) - EAt . In other words, we arrive at equation (5):

(5)

where dk and w are defined as before.

As before, we break up PRGk into its constituent parts: PRGk = Lk + Hk + EDk .

Substituting into equation (5), we arrive at equation (6):

(6)

Then, ERk = max {0, PRGk–EAk-1•(1+ik)}= max {0, Lk+Hk+EDk–EAk-1•(1+ik)}

Given that any existing experience account is paid at the time of recapture, we know that EAw = 0.

Thus,

PV ERd

dER

d

dPRG EAt

k

tk t

w

kk

tk t

w

k t( ) = = −= + = +∑ ∑

1 1

PV ERd

dL H ED EAt

k

tk k k

k t

w

t( ) ( )= + + −•

= +∑

1

EA EAd

dtEA

d

d

EAd

dEA

d

d

d

dEA i EA

t tt

ww

t

kk

tk

k

tk t

w

k

tk t

w

k k k

= −

= −

= + −

+ − •+ −

•

− •−

•

= +

= +

• − •

∑

∑

( )( )

( )

( ( ) )

1 11 1

11

1

1

1 1

continued on page 10


Substituting into equation (6) on the previous page, we get equation (7), for any point in time t:

Finally, when considering the present value of future cash settlements at any point in time t, wearrive at equation (8):

In other words, with a high degree of likeli-hood, the only items that matter in terms of thecash settlements is the change in the experienceaccount balance, with interest and fees on thatexperience account.

What we then see is:

1. There is an embedded derivative in boththe liabilities (i.e., within the preliminary rein-surance gain) as well as in the experiencerefunds.

2. On a present value basis, with a highdegree of likelihood, these embedded derivativecash flows precisely cancel each other out.

We therefore conclude that, taking intoaccount both embedded derivatives, the marketvalue of the single, net embedded derivative is 0at any given point in time for cash financialreinsurance transactions, as well.

Conclusion

Financial reinsurance transactions are struc-tured such that, with a high degree oflikelihood, the reinsurer will not experience amaterial, long-term economic loss. This leads tonot one, but two embedded derivatives, onewithin the funds-withheld asset and the otherwithin the experience refund provision. Thehigh likelihood of no long-term economic lossnecessarily implies that the net of these twoembedded derivatives will be equal to zero at alltimes.??


PV ERd

dL H ED EA

d

dL H ED

d

dEA i EA

d

dL H ED EA i

tk

tk k k t

k t

w

k

tk k k

k t

wk

tk t

w

k k k

k

tk k k

k t

w

k k

( ) ( )

( ) ( ( ) )

( ( )

= + + −

= + + − + −

= + + − +

•

= +

•

= + = +

• − •

•

= +

− •

∑

∑ ∑

∑

1

1 1

1

1

1

1

1 ++ EAk)

PV CS PV PRG ER

PV PRG PV ER

d

dL H ED

d

dL H ED EA i EA

d

dEA i EA

t t

t t

k

tk k k

k

tk k k k k k

k t

w

k t

w

k

tk t

w

k k k

( ) ( )

( ) ( )

( ) ( ( ) )

( ( ) )

= −= −

= + + − + + − + +

= + −

• • − •

= += +

•

= +

− •

∑∑

∑

1

11

1

1

1

1

Larry Carson is vice

president and actu-

ary RGA Reinsurance

Company. He can be

reached at

[email protected].

Embedded Derivatives... • from page 9

ance enterprises. Most top-tier companies havebeen downgraded by ratings agencies, manybecause of capital problems caused by otherlines of business.

The result of all of this consolidation hasbeen a change of direction in life reinsurancerates. Rates had consistently fallen for adecade, but new rates quoted have leveled offand in some cases have increased. Withconsolidation, finding a diversified pool isincreasingly difficult, and capacity is limited.Rates for group insurance, LTC and accidentand health have increased, and betweenconsolidation and the impact of 9/11, therates for catastrophe reinsurance coveragehave skyrocketed.

The rate increases are particularly hardfor today’s direct writers to deal with. Overthe last decade, ceding companies havebecome addicted to reinsurance. In 1993, only15 percent of new life insurance was rein-sured, but by 2000, the rate had reached 62percent, and in 2002 it was 61 percent. Manyof the direct writers have changed theirstrategy to focus more on accumulation prod-

ucts than protection products, leaving thereinsurers to manage the mortality risks andthe big reserves. “Ceding companies areaddicted, and there is no turning back,” saidMr. Clark.

Hank Ramsey suggested several ways thatthe ceding companies will likely respond. Mr.Ramsey, a vice president and actuary atPrudential Financial, said that companieswill likely retreat where they can from thehardening market. If XXX relief is only avail-able from a few reinsurers, and the price is asmuch as 50 percent higher than last year,then companies will “lose ground” every timethey renegotiate terms, and so will not re-bidas often. Companies may also considerretaining a larger percentage of each risk ifthe rates are not as attractive. As this higherretention leads to more earnings volatility,ceding companies may look more aggressivelyfor stop- loss programs. The t ighteningmarket for reinsurance will also result incompanies being less active and aggressive inthe term life market.

In addit ion to the issue of rates, Mr.



Implications of a Consolidated Market • continued from cover

1997 Rankings1. ERC2. Security Life3. RGA4. Transamerica5. Swiss Re6. Life Re7. Lincoln Re8. Phoenix Home9. Guardian10. Allianz Re11. American United Life12. Cologne13. BMA14. Manulife15. Munich Re16. CAN

2002 Rankings1. Swiss Re2. ING (Security Life +

Reliastar)3. RGA4. Transamerica5. Munich Re6. BMA/Generali7. ERC8. Annuity & Life Re9. Allianz Re10. Scottish Re11. Canada Life

2004 Rankings ? 1. Swiss Re2. ING (Security Life +

Reliastar)3. RGA4. Transamerica5. Munich Re6. BMA/Generali

10. Scottish Re11. Canada Life

Who Controls 90 Percent of the Reinsurance Market?

Source: SOA survey conducted by Munich Re

Hank Ramsey is vice

president and actu-

ary at Prudential

Insurance Company

of America in

Newark, NJ. He is

also a new member of

the Reinsurance

Section Council.

Hank may be reached

at hank.ramsey@

prudential.com.


Ramsey suggested that ceding companiesshould also be worried about the results ofconsolidation on counterparty risk. A lessdiversified pool of reinsurers means moreconcentrated counterparty risk. On the otherhand, reinsurers are using their newfoundleverage to push back on recapture triggers.[In December, Swiss Re announced a newglobal corporate policy against ratings trig-gers in their reinsurance agreements. ]Reinsurers are also obviously less comfort-able with change of control provisions.Finally, ceding companies are also concernedthat the reinsurers will offer less support forunderwriting manuals and intercompanymortality studies as they continue to squeezetheir expenses.

Mr. Ramsey sees some hope for increasedsupply of reinsurance in the future. Theincreased capacity may come from traditionalsources, as P&C reinsurers look to expandtheir operations, or it may come from unex-pected sources. For example, investmentbankers are aggressively seeking new securi-t ization transactions, which could addsignificantly to the capital capacity of lifereinsurers. However, securitization transac-tions to date have had some signif icantdownsides. They are generally more expen-sive than the more traditional Letter ofCredit approach, and they are only appropri-

ate for very large transactions.Paul Schuster, executive vice president at

RGA Re, said that the market today is “allabout profitability and capacity.” The resultof the price war in term reinsurance in recentyears is that all of the profits have been“squeezed out.” One response by reinsurershas been tighter contract terms. Anotherresponse is a hesitation to accept new kindsof risks. He views universal life policies withsecondary guarantees as “flawed products.”Ceding companies will either pay more toreinsure these products or will have to dobusiness with second-t ier reinsurers.Reinsurers are also requiring a higher stan-dard of financial reporting. He suggested thatfor ceding companies fast and accuratereporting of reinsurance transactions may bea competitive advantage in the future.

Mr. Schuster sees the industry’s need tofund XXX and AXXX reserves as the biggestchallenge. He estimated that the need will be$100 billion in seven years, but the bankLetters of Credit total only about $25 billiontoday. He asked how we will meet the $75million gap. He sees securitization transac-tions as the most likely factor to expand themarket in the future. But Mr. Clarksuggested that the growth in securitizationtransactions will be slow. “Investors don’t likerisks they don’t understand,” he said.??

Background

The Financial Accounting Standards Board(FASB) and the International AccountingStandards Board (IASB) issued “The NorwalkAgreement” in October 2002, a memorandumof understanding that reaffirmed theircommitment to develop a single set of high-quality accounting standards. At that meetingthe two standard setters agreed to place ahigh priority on three steps toward achievingthat goal:

1. Reduce, through a joint short-term proj-ect, (which is now mostly completed) thedifferences between U.S. GAAP and IASBstandards in certain areas not already beingaddressed by major projects.

2. Remove other differences through thecoordination of future work programs andcontinued progress on the joint projectsalready underway.

3. Encourage further coordination of theseparate activities of their two interpretivebodies.

International Financial Reporting Standards and Insuranceby Sam Gutterman

Implications of a Consolidated Market • from page 11

Of the projects being led by the IASB andmonitored by the FASB, the most relevant toreaders of this article is the development of anew international financial reporting standard(IFRS) for insurance contracts. This effort isthe culmination of work by the InsuranceSteering Committee of the IASB’s predecessor,the International Accounting StandardsCommittee (IASC), which began its delibera-tions in 19971. Extensive 1discussions havebeen held with many industry stakeholders,including the actuarial profession (led by theInternational Actuarial Association, alongwith many national actuarial organizationssuch as the AAA and the CIA). You may recog-nize various aspects of this discussion heldfrom various papers and presentations madeon fair value accounting over the last couple ofyears.

The current set of IFRSs does not containspecific accounting guidance for insurancecontracts. In fact, insurance contracts areexplicitly scoped out of a number of thementirely. Meanwhile, current U.S. GAAP insur-ance standards have primarily been developedover the past several decades under a match-ing (revenue and costs) framework whichtends to be income statement driven, althoughthe specific approach taken varies by type ofcontract. This is inconsistent in several waysfrom that being pioneered by the IASB’srevised accounting framework, which ismoving more toward a balance sheet orienta-tion.

Due to the complexity of the issues involved(principally whether “fair value” should beused to measure liabilities for insurancecontracts, and if so how to measure such avalue), the fact that the EU and Australiaissued directives that IAS standards wouldapply for all listed companies by 2005, and theresulting lengthy timeline needed to agree onthe standards needed for implementation, theproject was divided into two parts:

• Phase I – to provide initial guidance andfacilitate consistent compliance with IFRS forEuropean and other countries adopting IFRSin 2005 and

• Phase II – to incorporate in a comprehen-sive manner the more difficult recognition and

measurement concepts a couple of years later.The IASB Exposure Draft 5 Insurance

Contracts (ED 5) was issued in June 2003 aspart of phase I and was available for publiccomment until October. The IASB receivedmore than 130 comments on it. Then throughJanuary, the IASB Board discussed severalimportant and controversial issues relating toit. The newly named International FinancialReporting Standard 4 (IFRS 4), expected to bemade available in mid-March 2004 is theoutcome of these discussions. It is intended toserve as a bridge to phase II that will allowinsurers to continue most of their currentaccounting for insurance contract liabilitiesuntil the difficult issues in this area involvedare more fully addressed. At the same time, iteliminates certain “low-hanging accountingfruit” that shouldn’t require significantresources to change, such as European stabi-lization and Japanese catastrophe reservesthat are inconsistent with the IASB frame-work, which contains the basic concept underwhich the IASB’s standards are based, whileat the same time attempting to minimizeother significant deviations from the frame-work. The IASB Board is expected to begindiscussion of phase II issues in June of 2004.

The objective of this article is to provide thereader a basic understanding of what hashappened to date on phase I of this project,with a brief introduction to some of the keyissues that will be addressed in phase II.Please note that because some of the rules arenecessarily complex and this article was writ-ten prior to publication and implementation ofthese standards, the description provided maynot be completely consistent with practice aswill be applied. In addition, it does not coveractuarial standards that are currently beingdeveloped by the IAA.

IFRS 4

IFRS 4 (phase I) is the result of a series ofcompromises, adopted now primarily to satisfythe European Union (EU)’s requirement tomove to IASB’s standards in 2005, while at thesame time not creating the need for expensivesystems changes that might have to be



International Financial Reporting Standards and Insurance

1IFRS is the name for International Financial Reporting Standards, issued by the IASB. Under its predecessor these were referred to as IAS, orInternational Accounting Standards.

changed again when phase II is adopted. As aresult, several issues addressed here willlikely be revisited in the next year or year anda half until phase II is completed.

The following are some of the most signifi-cant issues addressed during the process ofdeveloping IFRS 4 and some of its key currentrequirements:

Insurance contract focus. The new insur-ance contract standard primarily addressesfinancial reporting for insurance contractsrather than for insurance companies, althoughit does incorporate certain requirements forcompany disclosures as well. In addition,phase I of the project also includes certainchanges to other financial reporting stan-dards, including IAS 32 and 39 (the twostandards dealing withfinancial instruments asboth assets and liabilities)and IAS 18 (the standarddealing with revenue,including those for servicecontracts).

Product classification.The insurance accountingapproaches used in U.S.GAAP differ depending onwhat type of product isinvolved. These categoriesinclude short-duration FASBStatement No. 60 (FAS 60),long-duration FAS 60, limited pay FAS 97,universal l i fe-type FAS 97, investmentcontract FAS 97 and mutual company FAS 120contracts. Similarly, at least through IFRS 4,the product category in which a contract isclassified will determine what measurementmethod should be used, although all aresubject to a liability adequacy test (see below).These categories are:

1. Insurance contracts. These are primarilyaccounted for by local GAAP rules (that is, ifU.S. GAAP is currently used, then it will beable to continue to be applied through the lifeof phase I), with some exceptions indicatedbelow.

2. Investment contracts (i.e., financial liabil-ities measured according to IAS 32 / IAS 39). Acompany is given a choice between the use ofan amortized cost or fair value method,

although limited guidance is currently avail-able regarding these methods.

3. Investment contracts with discretionaryparticipation features. This is a new categoryconsisting of various participating contracts.Local GAAP can be used, though they aresubject to a minimum value which is based onthe investment only (type 2) contract. Theseare not particularly common in the UnitedStates, but can constitute a significant percentof business inforce of insurers in countriessuch as France and Germany.

4. Service features. In accordance with IAS 18, if a class 2 or 3 contract has servicefeatures (e.g., variable or unit-linked productswith respect to assets managed), then adeferred acquisition cost (DAC) asset can be

established, but limited toincremental or marginalcosts, then subject to amorti-zation consistent with therevenue recognized.

The measurement meth-ods used can be changed(although somewhat complexcriteria must be met), butonly if they represent animprovement, that is, a movetoward a fair value-basedsystem, e.g., a move fromundiscounted to discountedliabilities.

Insurance definition. Due to potentiallysignificant differences in values between themethods applicable to insurance contracts andinvestment type contracts, the definition of aninsurance contract is quite important. Themost important distinction is between a finan-cial instrument (type 2 above) and aninsurance contract (type 1 above). The IASBhas attempted to categorize as many contractsas possible as insurance, in order to reduce thecomputational (systems) changes required inphase I.

IFRS 4 defines an insurance contract as a“contract under which one party (the insurer)accepts significant insurance risk fromanother party (the policyholder) by agreeing tocompensate the policyholder if a specifieduncertain future event (the insured event)adversely affects the policyholder or other


If a contract isdetermined to be

an insurancecontract, insurers

will apply theircurrent accounting

standards …

International Financial Reporting Standards and Insurance • from page 13

beneficiary.”To qualify as insurance, at least one of the

following uncertainties must be present: (1)whether an insured event will occur, (2) whenit will occur or (3) how much will be paid.Insurance thus includes retroactive reinsur-ance, in which the insured event would be theinsurance payment and not the original loss,but would not include most forms of financialreinsurance, even though the timing ofpayment is not certain. A key distinction isbetween insurance and financial risk, theformer of which requires some adverse conse-quences to the insured. However, the keyconcept underlying this definition is whether acontract has significant insurance risk, inwhich an insured event could cause an insurerto pay significant additional benefits in anyscenario, excluding any scenarios with nodiscernible effect on the economics of thetransaction. Many annuities will contain suchrisk if they include a minimum guaranteedannuitization benefit, although they will be aninvestment contract during their accumulatephase if there is no guarantee of rates onannuitization and no significant minimumdeath benefit. If a contract is determined to bean insurance contract at issue, it will continueto be considered an insurance contract;conversely, if a contract is determined to nothave sufficient insurance risk at issue, it canbe reclassified as an insurance contract at alater time.

While this definition will most likely resultin limited categorization differences from U.S.GAAP for U.S. products, some insurancecompany contracts will certainly not containsufficiently “significant insurance risk” underthe above definition (particularly manypension contracts and group contracts with acomplete experience refund, as well as finan-cial reinsurance), and will be subject to theIASB financial instrument / investmentcontract standards, IAS 32 and 39.

Like U.S. GAAP, no formula will be providedto measure “significant,” but doubtless somead hoc benchmark(s) may be developed inpractice, although it is highly doubtful that itwill be like the informal 10 percent chance of a10 percent loss rule.

What to do with insurance contracts. Ifa contract is determined to be an insurancecontract, insurers will apply their currentaccounting standards to insurance contractsuntil phase II is adopted. Some of the prac-tices specifically allowed in IFRS 4 butexpected to be eliminated in phase II that arecurrently followed by U.S. insurers are:

• Measuring insurance property/casualtyloss reserves on an undiscounted basis. Inphase II, it is likely that these liabilities willbe discounted, with an as yet undefinedadjustment for risk (often referred to as a“market value margin,” reflecting the market’scurrent appetite for risk)

• Reflecting future investments margins inthe measurement of insurance liabilities by i)using the estimated return on assets expectedto be held as a discount rate or ii) projectingthe yield on those assets at an assumed rate ofreturn, discounting the projected returns at adifferent rate and incorporating the result inthe measurement of the liability. This prima-rily affects non-variable l i fe insurancecontracts with a savings element. The inabilityto reflect such margins, without other offset-ting approaches, could lead in some cases torecognition of a loss at issue.

Liability adequacy test. Many of thetemporary compromises were made by theIASB assuming that a rigorous liabilityadequacy test would be applied (this is a newterm, but in concept it is similar to a lossrecognition test). This test must considercurrent estimates of all future cash flows froma contract, including embedded options andguarantees. If it fails, then an additionalliability is required with the resultant lossrecognized as a loss on the income statement.If current accounting policies do not include aliability adequacy test that meets the require-ments, then an IASB test, given in IAS 37,Provisions, Contingent Liabilities andContingent Assets, has to be applied.

Unbundling. If a contract is classified asan insurance contract, unbundling could berequired; that is, different accountingapproaches could apply to its components ifsufficiently different. If both an insurance anda deposit component is present and the depositelement can be measured separately and regu-




lar accounting policies (in the United States,U.S. GAAP) do not require it to be recognized,unbundling should occur. An example is if acedant receives compensation for losses from areinsurer, but the contract obliges the cedantto repay that compensation in future years,the obligation would be deemed to have arisenfrom a deposit component. Also, if the account-ing policies used permit compensation to berecognized as income without recognizing aliability, unbundling is required.

What to do with an investmentcontract. If an insurer’s contract does nothave significant insurance risk, it is thensubject to the requirements of the recentlyrevised IAS 32 and 39. Under these standards,such a contract will be carried at either its fairvalue or its amortized cost, as elected by theinsurer prior to issue. Although a completedescription of these standards is far beyondthe scope of this article, the following brieflydescribes some of their key provisions.

• Amortized cost. The interest method isbasically used, which solves for the interestrate needed to mature the contract.

• Fair value. What constitutes a fair valuefor these contracts is not yet clearly defined.Concepts such as deposit floor (the net liabil-ity cannot be valued for less than amount thatthe policyholders can demand), own creditstanding (reduction in the liability to reflectthe credit risk of the company), and treatmentof possible future premiums in a flexiblepremium contract may require IAA guidance.

Measurement inconsistency. One of themost controversial issues in the run-up toIFRS 4 has been the potentially misleadingfinancial information that can be generated byassets and liabilities measured on inconsistentbases. Consistency of measurement has been along-standing principle of the IAA in theseven-year-long discussion regarding thedevelopment of international accounting stan-dards. In addition, it was the topic of a recentjoint research project between the IAA and theAmerican Council of Life Insurance (ACLI),showing that income statements would beinconsistent with economic reality wheninconsistent measurement is followed as inter-est rates vary.

Volatile results and results inconsistent

with true underlying performance andeconomic reality will be reported for many lifeinsurance companies that apply IFRS 4 in2005, as they continue to follow what can beviewed as primarily an amortized costapproach (in other words, not being fullyresponsive to changes in the interest rateenvironment) in the measurement of theirliabilities, combined with fair value for most oftheir assets in accordance with IAS 39 (in the“available for sale” (AFS) category in whichchanges in fair value are reflected as anequity adjustment; at the January IASB Boardmeeting, the IASB agreed to expose forcomment the possibility of having changes inthe fair value of certain AFS assets flowthrough the income statement, but this mayprove to have limited value).

This mismatch in approach (i.e., particu-larly regarding the sensitivity to changes inthe interest rate environment) currentlyoccurs under U.S. GAAP due to the prevalenceof the use of the AFS asset category, used toavoid the strict tainting rules associated withclassifying these assets as “held-to-maturity,”but the impact is reduced by having assetvalue changes go through other comprehen-sive income (OCI) and the effect of shadowDAC reported separately. IFRS does notprovide for booking of fair value changesthrough OCI.

This concern arises particularly because ofpossible artificial losses that will be reportedin an increasing interest rate environmentthat might occur over the next few years,where the value of liabilities will remain rela-tively stable with a corresponding decline inasset values. Several possible solutions havebeen raised in the course of the IASB discus-sions over the past few months, including thepossibility of 1) relaxing the tainting rules inIAS 39 (which are generally consistent withFAS 115) to allow for easier classification ofassets as “held-to-maturity” which then can beheld at amortized cost, being potentially moreconsistent with changes in some U.S. GAAPnet liability values, 2) permitting the use of aseparate asset category, “assets backing insur-ance liabilities,” that would permit assets to beheld at amortized cost consistent with liabilitymeasurement, 3) allowing for the unlocking of



the discount rate used to calculate insuranceliabilities by block of business, or 4) providingfor some type of shadow accounting to reflectunrealized gains, similar to shadow account-ing in U.S. GAAP.

In IFRS 4, the IASB will not permit anyasset-based solution to this problem, althoughthe possible changes in reporting for AFSassets in IAS 39 can be viewed as a solution aslong as liability measurement is based oncurrent interest rates. However, the IASB willpermit a liability-based solution (no. 3 or no. 4above). The problem with no. 3 is that in manycases, it would require asignificantly revised valua-tion method which could bequite costly to implement. Toaccommodate this difficultinterest-sensitive liabilityapproach which few willutilize, the IASB Board maymake the AFS changementioned above. A fewEuropean insurers areexploring the shadow-accounting approach, but itis too early to tell how popu-lar this method mightbecome.

Embedded derivatives. IAS 39 alreadyrequires that some embedded derivativesshould be separately valued from their hostcontract, with the effect of a change in fairvalues flowing through income. IFRS 4exempts both an embedded derivative thatwould be an insurance contract if offered sepa-rately and surrender options in investmentcontracts with discretionary features andinsurance contracts. Nevertheless, expectedcash flows from these embedded derivativesshould be reflected in the liability adequacytests.

Disclosure. An insurer will be required toprovide information to understand theamounts in the financial statements that arisefrom insurance contracts and the amount,timing and uncertainty of their future cashflows. This will include information regardingthe effect on profit and loss from sensitivitytests involving material risk variables, claimreserve development, major assumptions, risk

management objectives, and on major risksand their concentration.

Reinsurance-specific items. In general,reinsurance accounting under ED5 appliesU.S. GAAP FAS 113 concepts. The followingspecific rules are included:

• No netting permitted. Gross and net ofceded reinsurance values will have to bereported separately.

• Profit/loss at issue disclosure. Althoughearlier drafts would not have permitted aprofit at issue, IFRS 4 allows a profit to berecorded at treaty issue, as long as this

amount is disclosed in thefinancial statement foot-notes. In addition, if theprofit/loss is deferred, theamortization of this amountand outstanding balance arealso to be disclosed.

• Reinsurance ceded hair-cut. Less than fullrecognition of a ceded rein-surance asset could berequired, reflecting thecredit standing of the rein-surer used. At the time thatthis article was written, itwas unclear what approach

would be taken to measure any such reductionin reinsurance credit.

This only touches the surface of a newsystem of financial reporting of insurancecontracts. This article has not addressed manyof what may be significant details that canaffect a particular company’s reserves. So,when IFRS 4 is published in mid-March,please look through the details. In addition,during the course of implementation, anumber of issues will likely arise that wereunanticipated when the standard was written.

The next stage

Many very difficult issues will be addressedover the next year and a half in the course ofcompletion of phase II of this project. Becauseof the difficult conceptual issues involved, it isuncertain how long the development of phaseII will take, although it will most likely not beadopted until at least late 2005, for possible


… some embeddedderivatives should

be separatelyvalued from their

host contract, withthe effect of achange in fairvalues flowing

through income.





implementation in 2008. A primary reason forthe long timeframe is the IASB has found thatchanges in accounting approaches for insur-ance contracts can have serious impacts onmany other fields of accounting as well. Forexample, changes made in IFRS 4 in the defi-nition of DAC for service contracts to includeall marginal acquisition expenses may have anunexpected effect on certain bank products.

A brief overview of some of the most impor-tant and controversial issues for phase IIinclude:

Fair value. What is the fair value of aninsurance contract? A key aspect of this ques-tion is whether a “pure” fair value definitionor a variation will be used (of which there aremany possible ones). For example, should fairvalue be the entry value, (the original pricecharged) or the current estimated exit orpurely prospective value? If it’s the latter, willa profit be allowed to be reported at issue?Many observers believe some form of fairvalue will be used in phase II, as the IASB ismoving in that direction in many areas.

However, Tom Jones, vice-chairman of theIASB, indicated in September that a fair valueapproach was not a “fait accompli,” and thatthe Board would keep an open mind in itsupcoming discussions regarding the bestapproach to take. In fact, if entry values areused, there may be little difference betweenfair value and a FAS 60 amortized costapproach, possibly even with a DAC and fairvalue. This debate will continue over theremainder of this IASB project.

Asset / liability measurement mismatch.Although there was a significant effort overthe last several months to arrive at a consen-sus solution, no resolution equally acceptableto all was reached in IFRS 4. Because it isunlikely that all assets will be valued on a fairvalue basis with changes affecting the incomestatement by the time phase II is imple-mented, the effect of measuring assets andliabilities on an inconsistent basis will likelyremain an issue for phase II discussions.

Unbundling. Certainly unbundling orbifurcation will be a topic that will beaddressed, even whether a traditional wholelife contract should somehow be split into itssavings and risk elements. Many observersbelieve that this is needed to provide compara-

ble values with other financial services firmsselling deposit or balance type financial prod-ucts, and that premiums for the savingselement should not be counted as revenue.

Embedded derivatives, options andguarantees. It is likely that more of these willbe required to be valued through eitherstochastic methods or demonstrated equiva-lents. This may present a challenge to manyactuaries who will have to be able to applygenerally acceptable methodology for themeasurement of the fair value of many ofthese benefits.

Loss reserves. Germane to P&C insurance,it seems that the use of discounted lossreserves is an almost foregone conclusion.However, probably more important is how thecorresponding risk margins (generally referredto as market value margins = risk marginsthat reflect the market’s perception of therisk) are determined for these liabilities.

Renewal premiums. Because renewalpremiums are not under the control of theinsurer (i.e., policyholders aren’t obligated topay them), they may not be acceptable to bereflected as an asset or anticipated beforecollected. A new approach may be required toavoid reflecting these in current measure-ment, although this might not provide acomplete picture of the structure of manyinsurance contracts.

Future investment margins. When toreflect expected profits from future interestearnings will be addressed. Currently theIASB Board is opposed to reflecting thesebefore the corresponding interest is earned. Asa result, initial losses on contract sales mayhave to be recognized for several insuranceproducts.

Minimum deposit floor. Many within theindustry believe that it is inappropriate toincorporate a minimum floor (cash surrendervalue or zero if there is none) to a liability,particularly if a DAC asset is not allowed. Thisis currently a constraint on recognition of theotherwise determined liability requirement forinvestment contracts.

Discount rate. Many believe that thecurrent risk-free rate should be used.However, difficulties can arise in certain cases,i.e., in countries in which there are no assetsto match the liabilities involved. Alternatives

Sam Gutterman,

FCA, FSA, MAAA is

director and consult-

ing actuary at

PRICEWATER-

HOUSECOOPERS

LLP. He can be

reached at

sam.gutterman@

us.pwc.com.


One of the life insurance industry’s resid-ual impacts of the tragic events of September11, 2001 was some marked changes in themarket for catastrophic reinsurance cover-ages. Reinsurers began re-evaluating theirrisk profiles, and consequently, changes inthe price and availability of this type ofcoverage were seen. The life insurance writ-ers of the world also began picking up on newsets of insurance vocabulary: terms like“terrorism exclusion” and “federal backstop”became increasingly used in reinsurancediscussions.

After 9/11, direct life insurance writersalso began evaluating different approaches tocatastrophic reinsurance coverages.Companies had choices to make as the cost ofsingle company coverages increased and cata-strophic pool arrangements were changing. Insome cases, many companies found them-selves in a posit ion where maximumexposure limits in catastrophic pools rose tonearly four times their original levels prior to9/11. At the same time, some companiesfound themselves in the unenviable positionof being mixed in catastrophic pools withother insurers who may have much higherprobabilities of having a catastrophic claim.This turn of events found many companiesseeking alternative ways to obtain cata-strophic reinsurance coverage, and evencontemplating the idea of carrying no cover-age at all.

One of the ideas to arise out of this evolv-ing situation was the creation of a newcatastrophic reinsurance pool arrangementfor the life insurance industry. This new pool,the Shared Adverse Fluctuation ExperiencePool Agreement (the “SAFE Pool” ) , wasdesigned to create catastrophic reinsurancecoverage for companies with a low concentra-t ion of l i fe insurance r isk in majormetropolitan areas.

The SAFE Pool began operating on July 1,2003 with American Farm Bureau InsuranceServices serving as the pool administrator.

Twelve life companies currently are poolmembers and are contributing approximately$70 billion of mortality risk. With the initiallimit of liability set at $0.10 per $1,000 of inforce, the initial term provided maximumrecovery for each company and for the pool intotal of $7 million. Catastrophic claims canbe filed by a pool member if the memberexperiences any type of incident that resultsin at least four insured deaths. As with othercatastrophic pools, no risk premiums are paidand all claims against the pool are fundedthrough assessments against member compa-nies. Claim payments are paid according tothe percentage of in force each membercontributes to the pool. Annual administra-tive service fees in 2003 ranged between$3,000 and $4,000 per member depending onthe size of the member ’s in force. Newentrants can be added at the beginning ofany calendar quarter.

The pool leverages off the idea of “cata-strophic underwriting” commonly seen insingle company coverages to ensure the poolonly accepts members with similar riskprofiles. In-force listings by zip code areanalyzed to determine the amount of riskconcentrated in large urban areas, and ques-tions regarding life insurance risk outsidethe United States and Canada are commonlyasked. While the definition of a “preferredcatastrophic risk” is hard to define, poolmembers are at least ensured that thecompanies in the SAFE Pool have similarrisk characteristics.

At the SAFE Pool ’s annual AdvisoryCommittee meeting in October, pool membersdiscussed future changes that could furtherassist the catastrophic reinsurance needs ofmember companies. Pool members discussedraising the maximum recovery limit to $0.15per $1,000 in the future and also purchasinga second layer of coverage to expand the totalcoverage to $20 million. The cost of the addi-tional layer has been seen to be a more cost


SAFE Pool Provides New CatastrophicCoverage Alternative for Life Insurance Industryby R. Dale Hall

R. Dale Hall is chief

life and annuity actu-

ary at COUNTRY

INSURANCE and

Financial Services.

He can be reached at

dale.hall@countryfi-

nancial.com.


SAFE Pool... • from page 19

effective approach due to the large diversifi-cation of risk of the member companies andthe ability to share the cost of the coverage.

While everyone in the industry hopes theevents of September 11, 2001 are a one-timeoccurrence, it’s encouraging to see that new

ways to deal with the risk of catastrophicevents are evolving. The SAFE Pool appearsto be among these new ideas that can helpprovide stability to the financial strength ofits members even if future catastrophicevents were to occur.??

For a variety of reasons, insurancecompanies significantly expandedtheir use of life reinsurance through-

out the late '90s and early '00s. This hasmeant that volumes being ceded to the rein-surance market have continued to expand(after a brief respite in 2001) through quotashare opportunities with direct insurerskeeping only a portion of their publishedretention. The drive for growth and volumeled the reinsurers to try to offer more per-lifecapacity to the market by looking for increas-ing automatic binding l imits and jumbolimits from their retrocessionaires. Throughthe late '90s, most of the life retrocessionoutlets including the two ful l service,professional retrocessionaires (ManulifeReinsurance and Sun Life Reinsurance) wereable to offer greater automatic binding limitsand jumbo limits to service their life reinsur-ance clients who, in turn, offered higherlimits to their direct writers. Direct writershad access to more than 25 life reinsurersactive in the U.S. market and reinsurers andretrocessionaires typically also had access toEuropean and Asian reinsurers not active inthe U.S. market, who were willing to provideretrocession capacity on U.S. lives. So whathas changed? I wil l try to give you theperspective of a company at the top of thecapacity pyramid.

Clearly, the movement to quota-share rein-surance meant that direct writers were

retaining less on a per-life capacity basis.Massive U.S. life reinsurer consolidation(Lincoln Re, AUL Re, Phoenix Re, CNA Re,Cigna Re, Allianz Re, Life Re, to name a few)has resulted in less choice for the Direct writ-ers. It has also resulted in the loss of per-lifecapacity as the acquiring reinsurers havenot, generally, increased their retentionssufficiently to make up for the loss of capac-ity owing to the acquisitions. This problemwill only be further exacerbated by ERC’srecent announcement of their withdrawalfrom accepting new business going forward.

At the same time, many of the retrocessionoutlets for U.S. lives, smaller European rein-surers with little or no active U.S. operations,have also been acquired by the larger multi-lined and multinational reinsurers who arealready active in the U.S. market. Finally,some of same smaller reinsurers have beenhurt by large early duration claims thataggregated from their various retrocessionrelationships to a level that they wereuncomfortable with, forcing many of theseremaining companies to either stop acceptingretrocession on U.S. lives or severly reducetheir offered capacity.

I estimate that all of the above factorshave resulted in a reduction of per-life capac-ity in the United States by more than $100million. Considering the market started withsomewhere between $225 and $300 million ofcapacity, this is a material reduction that is

Capacity in the U.S. Life Insurance Market – A View from the Top of the Pyramidby Michael DeKoning

rippling through the market.So how is this affecting the industry on a

day-to-day basis?Autobinding limits are under signifi-

cant pressure. For retrocessionaires, beingexcess of both direct writers’ and reinsurers’retentions means that typically only largepolicies see their way through to the retroces-sionaire. The reduction in the number ofreinsurers keeping a retention on the oneside coupled with the reduction in the retro-cessionaire’s own retro pool capacity, hasresulted in many retrocessionaires beingforced to reduce the automatic binding limitsthey offer to the reinsurers. This, in turn,affects the amount of capacity offered by thereinsurers to the direct writers.

Increased pressure on jumbo limits.Jumbo limits, typically defined as the amountof insurance in force and to be placed on a given individual at application time, are tools used by reinsurers and retro-cessionaires to control aggregation on verylarge policies. A policy on an applicant that isbelow the jumbo limit can typically be cededautomatically under a reinsurance treaty(assuming it meets the other automatic bind-ing criteria based on age, underwrit ingrating, etc). Back in the early '90s, jumbolimits typically were in the $20-25 millionrange. In the mid to late '90s, jumbo limitsexploded to $75 million and even, in limitedcircumstances, to unlimited amounts. For theretrocessionaires, the jumbo limit is probablythe most important risk aggregation manage-ment tool there is. For any large policy, therecould be mutiple direct writers and multiplereinsurers, but there are only a small numberof retrocessionaires. As the risk takers at thetop of the pyramid, retrocessionaires willinvariably see the same life from multiplereinsurance arrangements. For a $75 millionpolicy, for example, it is very likely that morethan $55 million of this policy will end up inthe retrocession market. The large retroces-sionaires, like Manulife and Sun, could easilyend up with $25-35 million of this one risk.As retrocession capacity dries up behind theactive U.S. retrocessionaires, it is clear thatcontrolling the amount of risk being cededautomatically on large policies is key to

controlling risk aggregation and retentionmanagement. While I do not see jumbo limitsreturning to their early '90s levels, I wouldexpect that jumbo limits will reduce over thecoming 12-24 months.

Data quality and lags. In addition to theaforementioned risk aggregation issuesfacing retrocessionaires, the other big issue isthe lag in receiving accurate and detailedreporting for retrocessionaires to performretention management. As the companies atthe end of the reporting chain, the typical lagfrom the time a retrocessionaire is bound(which is concurrent with the policy issuedate) to the time it actually received report-ing is 18 months. This lag can be as long as36 months due to late reporting, systemschanges by either the insurer or reinsurer (orboth!) or poor quality administration. Theability to effectively manage retention andover-retention situations as well as offerfacultative capacity is severely limited due tothe data problems in our business.

Higher retrocession costs. Due to all ofthe above risk and supply issues, the cost ofretrocession in the life market has beenunder s ignif icant pressure. The cavernbetween inward premiums and the outwardcost of ceding to non-U.S. retrocessionaires,still an important outlet for the U.S. reinsur-ers and retrocesionaires, has continued to


Capacity in the U.S. Life Insurance Market…


widen. U.S. retrocession volumes havedropped over the last few years primarily dueto the lower number of quota share opportu-nit ies of fered to retrocessionaires. Theamount of business assumed from high faceamount policies, however, has only increasedover the last number of years. Therefore, theproportion of business assumed by U.S. retro-cessionaires that come from very largepolicies continues to increase. This means,however, that the higher retro costs havebecome increasingly burdensome to the U.S.retrocessionaire.

So, what can be done to stem this loss ofper-life capacity? Be nice to your reinsurersand retrocessioniares.

A dramatic improvement in data flowis needed. We are all part of an industrythat does not, at its core, have a closelyfollowed data standard (there is a standardin place that is not well-adopted) nor does ithave a common methodology for passing databetween participants. Each insurer passes itsdata on to their reinsurers in their ownunique format who in turn must translatethat data, process i t and then pass on the relevant information to their retroces-sionaires. Can we not find a solution to thisproblem? The banks have done it, the invest-ment firms have done it, what makes ourbusiness so di f ferent? There have beennumerous attempts to standardize or evencreate a data hub for the passing of data

from insurers to reinsurers and ontoRetrocessionaires. I believe these types ofinitiatives need widespread industry supportand can only improve data flow, cash flow andrisk management. This will lead to greatercapacity being avai lable sooner to themarket.

Current market market conditionscould lead to insurers retaining more oftheir business. If market capacity continuesto shrink, LOC costs increase and the often-rumored hardening of life reinsurance ratescomes to pass, it is possible that some compa-nies could turn their backs on first-dollarquota share reinsurance and return to retain-ing more business. This would increase theamount of total per-life capacity in the over-all market.

Reduced collateralization require-ments for U.S. business. While not a largeimpediment to non-U.S. companies, I believethat reducing the collateralization require-ments for non-U.S. companies could increasethe number of companies willing to acceptU.S. risks.

Increased profitability of life reinsur-ers in isolation and relative to the P&Creinsurance market. Many U.S. life rein-surers are struggl ing to meet theirshareholder return requirements. On a U.S.GAAP basis, very few if any, are makingdouble-digit returns on their new business. Itshould come as no surprise, therefore, thatthere are few potential new entrants and/orfew looking to expand in this market fromtheir current position. This is especially truegiven the P&C reinsurance market iscurrently in the midst of a hard market, withpotential returns often quoted in the 20percent-plus range. I suggest you ask your-selves this one question: If you were going toinvest in the insurance or reinsurancemarket today, would you invest in the lifereinsurance market?

Increased use of facultative reinsur-ance due to lower automatic bindinglimits. Due to the aforementioned reductionin retrocessionaire and reinsurer automaticbinding limits, the inevitable result will begreater amounts of facultative business. This,from a retention risk management point ofview is a good thing, in that both reinsurers


Capacity in the U.S. Life Insurance Market… • from page 21

Mike DeKoning has

responsibility for life

retrocession, struc-

tured reinsurance

and accident reinsur-

ance at Manulife

Reinsurance, a divi-

sion of Manulife

Financial. Mike can

be reached at

mike_dekoning@man

ulife.com.

The recapture provision is a standardreinsurance provision found in practi-cal ly every reinsurance treaty.

Historically, reinsurance was ceded on anexcess basis (i.e. the amount reinsured wasequal to the face amount inexcess of the company’s reten-t ion schedule) . The overal lratio of the reinsurance amountceded compared to thecompany’s direct faceamount was relatively low.The main purpose of“excess reinsurance” was toenable the direct writer toretain as much faceamount as i t couldjustify and merely cedethe amounts which it fe lt wasexcessive relative to its surplus,earnings or other financial criteria. Asexperience unfolded, the direct writer wasnot especially concerned about the relation-ship between the mortality experience of thereinsured business and the reinsurancepremium. (As we will soon discuss this iscertainly not the case under the more

recently utilized first dollar quota share rein-surance). The recapture provision was alogical, reasonable and benign provision thatpermitted the ceding company (i.e. gave it the

option) to increase its reten-tion limits on its in-forcebusiness (i.e. take back orrecapture some of thereinsured business) if itincreased its retentionlimits on new business.

If the increased reten-tion limit exceeds the face

amount of the policy rein-sured, then that policy willbe ful ly recaptured.Otherwise, it will be recap-tured only to the extent of the

increase in retention. Therecapture provision typically

has requirements such as arecapture (waiting) period (typically 10

years) as well as advanced notification ofintent to recapture. Some recapture provi-s ions require that the ceding companyimplement a recapture program within a


Capacity in the U.S. Life Insurance Market…

THE RECAPTURE PROVISION IS IT UP TO DATE?by Larry Warren


and retrocessionaires have the ability tobetter manage and control over-retention situ-ations, thereby allowing them to offer morecapacity without having to “hold any back.”That said, this will only become a factor oncethe industry has been able to address its dataissues and reduced the reporting lag to six tonine months.

I suspect the next few years are going to bevery interesting in the mortality risk market.The dynamics are very fluid, with significantopportunities for both improvement in marketefficiencies and risk management. That said, Ibelieve the next 12-24 months will also see

some interesting per-life capacity develop-ments that could drastical ly change theinsurer/reinsurer/retrocessionaire relation-ship. While I would not expect a return to the“strictly excess” and significantly limitedautomatic binding and jumbo limits that char-acterized the life reinsurance and retrocessionmarkets up to the mid-'90s, I believe that thetrend toward loosening these terms wil lreverse somewhat in the coming months.??

limited period after the effective date of aretention scale increase (otherwise, they willforfeit the right to recapture). There are alsoadministrative and other rules that have tobe followed. The purpose of the recaptureprovision is to give the direct writer theopportunity to take back the reinsured riskthat is no longer needed as a result of theincreased retention that it is now able toaccommodate. The recapture period must belong enough to give the reinsurer sufficienttime to earn its profit.

In a first dollar quota share arrangement,the reinsurers assume a fixed percentage ofthe face amount of each policy. For example,

the direct writer may retain 20 percent of each risk and cede 80 percent to one ormore reinsurers. First dollar quota sharereinsurance (80-90 percent is common) hasbecome quite prevalent in recent years andaccounts for a very significant percentage ofall reinsurance ceded. Many of these treatieshave similar if not identical recapture provi-sions as the “excess reinsurance” treaties.While the language in these provisions wasclear and appropriate for excess reinsurance,it is unclear and inappropriate for quotashare reinsurance and poses a very real riskfor disputes between the direct writer andreinsurer. There inevitably will be somedirect writers that, by the time the recaptureperiod is near completion, will recognize thatthe reinsurance premiums that they arepaying are greatly in excess of mortality

claims. As we will later show, it is the combi-nation of a strong and even perhapscompelling desire of the direct writer torecapture, coupled with this inappropriateand unclear language, which will spark majordisputes leading to arbitration and/or legalchallenges.

In sharp contrast to excess reinsurance,f irst dol lar quota share reinsurance isutilized for reasons basically unrelated to thedirect writer ’s retention scale, such aspredictable mortality costs (i.e. paying knownreinsurance premiums instead of unknownfuture mortality claims), stability of earn-ings, abi l i ty to of fer more competit iveproducts etc.

As a result of the fact that reinsurerscommonly build future mortality improve-ments into their pricing, coupled with thefact that projecting future mortality is an artas well as a science (i.e. determining whichmortality table has the appropriate slope forthe business being reinsured), it is not excep-tional to find reinsurers who will offer areinsurance premium rate scale lower thanthe ceding company’s pric ing mortal ityassumption. This lower premium wouldenable the direct writer to develop a morecompetitive product than it would be able tootherwise justify.

As mentioned earlier in our discussion ofexcess reinsurance, the “direct writer is notespecially concerned about the relationshipbetween the mortality experience of the rein-sured business and the reinsurancepremium.” This is because the direct writercould not prudently have kept a risk greaterthan its maximum retention scale. It simplyhad no choice but to reinsure the business.Furthermore, the volume of business rein-sured under excess reinsurance is typicallylow in relation to the total volume of directbusiness and is usually not of sufficient sizeto be statistically credible. As mentionedearlier in first dollar quota share arrange-ments, a very significant percentage of theface amount is typically reinsured (80-90percent is not exceptional), giving rise tohuge blocks of in-force business and often isof sufficient size to be statistically credible.In quota share arrangements, both the ceding


The Recapture Provision • from page 23

company and the reinsurer have a big inter-est in how the relationship between mortalityclaims and reinsurance premiums unfolds.

If mortality turns out to be significantlymore favorable than the direct writer hadcontemplated, the direct writer will makeevery attempt to recapture the reinsuredbusiness. In fact, I believe that it will not beuncommon for there to be situations wherethe direct writers will find themselves payingreinsurance premiums greatly in excess ofmortality costs. Let us look at the followingexamples below.

Example 1The direct writer, having

no credible mortality experi-ence (e.g. for a new productwith new risk classes or new underwrit ing guide-lines/requirements), makesan educated guess (based onsubjectivity and judgment)at what they think a reason-able mortality assumptionis. The reinsurers also haveno mortality experience onwhich to base their premi-ums. They similarly make aneducated guess based on the direct writer’smanagement team, distribution system,specific product, design, underwriting guide-lines, market segment, average face amount,etc. The direct writer then reinsures on a 90percent first dollar quota share basis with areinsurer or reinsurers whose YRT premiumsare lower than their mortality assumptions.They are initially quite pleased that they arelocking in higher profit margins throughreinsurance. After a few years elapse andcredible statistical experience emerges, thereinsurance premiums turn out to be consid-erably higher than the mortality claims. Thein-force business under this treaty (contain-ing several years of new issues) is now huge.The direct writer will be thinking, “if we wereonly able to recapture this business we willsave millions of dollars.” That is, they will behighly motivated to recapture the business.What recourse do they have? Exactly whatdoes the recapture provision permit them to do?

Example 2The direct writer has a reasonably good

idea of the mortality experience that theyhave had and their mortality assumption isbased on “accurate” mortal ity studiesrecently performed by the company. Thesemortality studies may even be statisticallycredible and based on the last three years ofexperience, which is ref lect ive of theircurrent underwriting guidelines/require-ments. Similar to Example 1, the directwriter after strenuous negotiations withseveral reinsurers, finally implements a firstdollar quota share arrangement with one or

more reinsurers, whoseYRT premium rates aresomewhat lower thantheir mortality assump-tions. This sounds toogood to be true as theywould be locking inhigher prof it marginsthrough reinsurance, andthis is even after sharingthe results of theirmortal ity study with the various reinsurers’bidding. As was the casein Example 1, after a few

years elapse it becomes quite apparent thatthe mortality claims are considerably lowerthan the reinsurance premiums. In thisexample, this result is from the fact thatdirect writers are not accustomed to buildingmortality improvements into pricing theirproducts since various regulatory require-ments such as self-support testing and policyillustrations usually prohibit it. Reinsurers,on the other hand, typically do factor mortal-ity improvements into their premium scales.Needless to say, there wil l be a certainpercentage of these quota-share arrange-ments where the annual mortal ityimprovements will turn out to be significant,giving rise to a greater and greater disparitybetween reinsurance premiums and mortalityclaims. That is, the aggressively pricing rein-surers who won the bid guessed correctly. Asin Example 1, this creates a situation wherethe reinsurance premiums eventually becomeconsiderably higher than mortality claims for




… reinsurance premiums begin tosignificantly exceedthe mortality claims.This may be quite a

surprise to the“astute” pricing

actuary.

a large in-force block of business and thedirect writer will be highly motivated torecapture the business. What recourse dothey have?Example 3

In this case, the direct writer’s pricingactuary is a l i tt le more astute than inExample 2 and takes pride in the mortalitystudies performed with his company’s newsophisticated mortality system. He uses themore “modern” 1990-95 select /ult imatemortality table (as opposed to the 1975-80select/ultimate mortality table) to develop hispricing mortality assumptions. He further-more has the reinsurers base their premiumson this table. He is perceptive and does infact realize that potential future mortalityimprovements are often recognized by thereinsurers and seeksout reinsurers with themost l iberal pric ingassumptions, includingan implicit aggressivemortality improvementassumption. He there-fore expects thereinsurance premiumsto be perhaps a fewpercent lower (e.g. 2.5percent) than his ownpricing mortal ityassumption. This iseven after allowing forthe fact that the rein-surer needs to cover itsexpenses and profit margin.

Once again, however, similar to the previ-ous examples, the reinsurance premiumsbegin to significantly exceed the mortalityclaims. This may be quite a surprise to the“astute” pricing actuary. However, this can infact happen when the reinsurance premiumsare expressed in terms of the 1990-95select/ultimate mortality table and yet thecompany’s mortality experience follows the1975-80 select/ultimate mortality table. Thissituation is shown in Exhibit 1 where Table 2(2.5 percent lower than Table 1) representsthe reduced reinsurance premium and Table3 represents actual mortal ity claims.Recognize the fact that Table 1 and Table 3

(based on the 1990-95 and 1975-80 mortalitytables respectively) were developed with scal-ing factors of 80 percent and 44.5 percentrespectively, making them equivalent over athree-year mortality study period. This equiv-alence can be seen by observing that the sumof the first three years for Table 1 and Table3 are each $2,320,000. Also in Exhibit 1, it isinteresting to observe in the last column“Excess Reinsurance Premium” that in theearly years (years two to four) the cedingcompany recognizes modest gains followed byever-increasing annual losses in the range of$1-2.9 million over the years 11-20 whichmay have been subject to recapture depend-ing upon the language in the treaty. Thereinsurance premiums are increasing at afaster rate than the mortality claims, because

the 1990-95 mortalitytable is steeper thanthe 1975-80 mortalitytable. As mentioned,the reinsurance premi-ums wil l begin tosignif icantly exceedmortality claims. Firstdol lar quota sharearrangements startedto rapidly gain in popu-larity in the mid to late'90s. Many of thesetreaties wil l soon benearing the end of their“10-year” recaptureperiod. As shown in the

above examples, there will very likely be astrong motivation on the part of some of thedirect writers to recapture their business.

In Example 1, due to the signif icantamount of judgment and subjectivity, theoutcome could very likely have been reversed.That is, mortality claims could have greatlyexceeded the reinsurance premiums as expe-rience unfolded. In Example 2, had themortality improvement not materialized, thesituation also would very likely be reversedwith the mortality claims exceeding the rein-surance premiums. In Example 3, there willin fact be cases where the mortality claimswill follow the slope of the 1990-95 mortalitytable and the reinsurance premiums will



In today’s environment, the ceding companynormally does due

diligence in the selection of their

reinsurers.



EXHIBIT 1DEMONSTRATION OF THE DISPARITY WHICH MAY ARISE BETWEEN

REINSURANCE PREMIUMS AND MORTALITY CLAIMS

TABLE 1 * TABLE 2** TABLE 3 *** TABLE 4

Reduced Excess Reinsurance Reinsurance ReinsurancePremium Premium Claims Premium

Year (= Table 2 - Table 3)

1 $550,000 536,250 $520,000 $16,250 2 780,000 760,500 770,000 $(9,500)3 990,000 965,250 1,030,000 $(64,750)4 1,190,000 1,160,250 1,220,000 $(59,750)5 1,440,000 1,404,000 1,390,000 $14,000

6 1,740,000 1,696,500 1,540,000 $156,500 7 2,120,000 2,067,000 1,690,000 $377,000 8 2,520,000 2,457,000 1,840,000 $617,000 9 2,900,000 2,827,500 2,030,000 $797,500

10 3,340,000 3,256,500 2,260,000 $996,500

11 3,740,000 3,646,500 2,580,000 $1,066,500 12 4,340,000 4,231,500 2,960,000 $1,271,500 13 5,020,000 4,894,500 3,440,000 $1,454,500 14 5,470,000 5,333,250 3,940,000 $1,393,250 15 6,010,000 5,859,750 4,460,000 $1,399,750

16 6,940,000 6,766,500 5,290,000 $1,476,500 17 7,860,000 7,663,500 5,860,000 $1,803,500 18 8,860,000 8,638,500 6,480,000 $2,158,500 19 9,980,000 9,730,500 7,150,000 $2,580,500 20 11,050,000 10,773,750 7,880,000 $2,893,750

* Represents 80% of the 1990-95 select/ultimate table based on mortality experience of the first3 policy years

** Table 2 is 97.5% of table 1*** Represents 44.5% of the 1975-80 select/ultimate table based on mortality experience of the

first 3 policy years

note: The mortality experience underlying these values was arbitrarily chosen to equal 80% ofthe 1990-95 select/ultimate table which is equivalent to 44.5% of the 1975-80 select/ulti-mate table.

For simplicity this exhibit is based on a single year of issue ($1 billion face amount) male issue age45 with zero lapses.


have been based on the 1975-80 mortalitytable. In these situations, the mortalityclaims will increase at a faster rate than thereinsurance premiums and will begin tosignificantly exceed them.

In all three newly defined “alternate”

examples above, in order to avoid significantlosses the reinsurers will desperately (due tothe large in-force block of quota share busi-ness) attempt to raise their rates especiallywhen the premium guarantee provision inthe treaty is unclear or ambiguous (as issometimes the case in YRT reinsurance).

It should now be apparent that both thereinsurer and the direct writer are taking bigrisks with first dollar quota share reinsur-ance. Depending upon the outcome, either thedirect writer or the reinsurer will have strongmotivation to take extreme measures toimprove their situation. From the directwriter’s perspective, as alluded to earlier,every attempt will be made to recapture theirbusiness. From the reinsurers’ perspective,every attempt will be made to raise premiumrates (on YRT reinsurance).

Reinsurers are nearly unanimous in theiropinion that no business under first dollar

quota share arrangements be eligible forrecapture. They properly recognize that therewould simply be too much selection againstthe reinsurer if recapture were permitted (i.e.if claims are very high, the direct writer willobviously preserve the reinsurance arrange-ment indefinitely, alternatively if claims arevery low, the direct writer will want to recap-ture). As previously mentioned, the recaptureprovision in most reinsurance treaties areunclear or ambiguous for first dollar quotashare arrangements.

For example, some treaties have no limita-tion at all regarding the business eligible forrecapture. They merely allude to a recaptureperiod (often shown on a separate schedulepage). Other treaties refer to the fact thatfacultative and reduced cessions are not eligi-ble for recapture, but never clearly identify ordefine quota share arrangements as reducedretention. Rather than define quota share asreduced retention and then let the cedingcompany deduce that it is not subject torecapture, the treaty language should clearlystate that the business ceded under this firstdollar quota share treaty is not eligible forrecapture. Treaty provisions are often silentas to whether an increase in the cedingcompany’s quota share retention from 10percent to 100 percent represents a trueincrease in retention scale or not. (Of course,the ceding company would assert that it is, tostrengthen its attempt to justify recapture).

Since it is typically the reinsurers’ intentthat quota share business not be subject torecapture, the treaty provision languagemust clearly and unambiguously state thisfact.

Until such time that the reinsurers reviseand clarify the recapture provisions in theirexisting treaties, we will find direct writersfalling into situations arising from the vari-ous examples previously discussed, who willbe compelled to focus on any ambiguous,unclear or vague treaty language. This focuswill enable them to justify recapturing theirbusiness in order to avoid significant losses.

In today ’s environment, the cedingcompany normally does due diligence in theselection of their reinsurers. This includesreviewing the reinsurers ’ rating agencyratings, risk-based capital ratios, financial



statements, etc. In order for the cedingcompany to protect itself in some future timeperiod when the reinsurer’s financial condi-t ion may have seriously eroded, it iscustomary to have a treaty provision (oftenreferred to as the “insolvency provision”)containing various triggering events forwhich the ceding company would have theright to recapture. It is not uncommon to findtriggering events such as:

(a) The reinsurer becomes insolvent,impaired or unable to pay debts

(b) The reinsurer is about to be liquidatedor dissolved

(c) The reinsurer experiences a significantrating downgrade from two or morerating agencies

(d) A significant reduction (50 percent ormore) in the reinsurer’s surplus or riskbased capital ratio

(e) etc.As was the case in our previous examples,

where the direct writer will make everyattempt to find loopholes or ambiguities inthe recapture provision in order to preventsignificant losses, the direct writer will alsoattempt to find loopholes or ambiguities inthis “insolvency provision.” For example, theterm “impaired” in (a) is not clearly definedor the “rating downgrade” in (c), which a rein-surer may experience could be for benignreasons but the ceding company will jump on

their opportunity to recapture.It should now be apparent that judgment

and subjectivity in the process of projectingfuture claims or reinsurance premiums play alarge role for both the direct writer and thereinsurer. This uncertainty inevitably leadsto winners and losers in this guessing gameof future mortality rates versus appropriatereinsurance premiums. The huge volume ofbusiness associated with an in-force block offirst dollar quota share reinsurance greatlymagnifies the loss to either party, compellingthe direct writer to attempt to recapture (oralternatively compelling the reinsurer underYRT reinsurance to raise rates). It should benoted that due to the 10-year recaptureprovisions common in automatic first dollarquota share pools, and given that the use ofquota share reinsurance began escalating in1995, we will begin to see attempted recap-ture become more of a reality beginning in2005.

The concepts addressed in this articleshould provide a wake-up call to both thedirect writer and the reinsurer to very care-fully scrutinize the recapture provisions (alsothe insolvency provision and the premiumguarantee provision) in their treaties andassure that it is clear, precise and up todate.??


continued on page 1


Larry Warren FSA,

MAAA is executive

vice president and

chief actuary of

National Benefit Life

(NBL) and is respon-

sible for

NBL/Primerica Life

Reinsurance. He

can be reached at

larry.warren@

primerica.com.

Volunteers Wanted!

Would you like to have an impact on the future direction of the Reinsurance Section? Submityour name for consideration for the 2004 Reinsurance Section Council slate. The Councilserves section members by sponsoring continuing education and providing information toassist members in their work in the reinsurance area. Any SOA Reinsurance member whowould like to be considered as a candidate for the section council should contact Mel Young [email protected] as soon as possible.

The Reinsurance Section is pleased to be sponsoring three sessions at both upcoming spring meet-ings. The Pension and Health Specialty track meeting is on May 19-21, 2004 at the Anaheim Hiltonin Anaheim, California and the Life Specialty track meeting is on June 14-15, 2004 at the SanAntonio Marriott River Center in San Antonio, Texas.

The sessions we will be sponsoring and their dates are as follows:

Anaheim Meeting (Wednesday, May 19 to Friday, May 21)Long-Term Care Reinsurance5/19, 10:30 am – 12:00 pm (Panel Discussion)What to assume? Long-term care insurance is a very new product. The emerging experience forcesdirect writers and reinsurers to constantly reassess assumptions, particularly persistency.

This session covers:• How reinsurers and direct writers respond to emerging experience• How they work together to modify contract terms or to take other actions to represent and protect

each party’s interest

This session is designed for attendees who have moderate experience with the subject.

Terminal Funding, The Longevity Risk5/19, 4:00 pm – 5:30 pm (Open Forum)In the past three years, stock market declines wreaked havoc on pension plan assets. As a result,many plans have terminated. The contracts covering the terminated vested pension obligations havea significant risk that the pensioners will outlive the underlying mortality table.This session addresses approaches to develop assumptions for reinsurance covering the underlyinglongevity risk.


HMO Reinsurance5/20, 8:30 am – 10:00 am (Open Forum)Where’s the capital? HMOs suffer from a shortage of capital more than the insurance industry ingeneral. State regulators and rating agencies are focused on these capital issues. This session covershow reinsurers provide capital solutions to HMO’s through both traditional and non-traditional rein-surance.



SPRING MEETING REINSURANCE SESSIONS by John Nigh


San Antonio Meeting (Monday, June 14 – Tuesday, June 15)How Creditworthy Is Your Reinsurer?6/14, 10:30 am – 12:00 pm (Panel Discussion)Significant consolidation of reinsurers, as well as direct writers, has occurred over the last decade.Reinsurers have capital demands not unlike those of direct writers. Reserve credits and reinsuranceamounts owed represent a significant asset on direct writers’ balance sheets.

This session discusses:• Means to ensure collectability when claims occur• What can be done when reinsurers encounter financial problems

This panel addresses these and other issues. This session is designed for attendees who have moder-ate experience with the subject.Where Is Your Reinsurance When You Need It?6/14, 2:00 pm – 3:30 pm (Debate)Denials of reinsurance coverage are increasing, resulting in increased arbitrations between directwriters and reinsurers. A variety of reasons exists for the increase in frequency of denials, includingcontract termination and exclusion of specified claims.

Join this lively debate as the point of view regarding the rights to deny claims is presented by arepresentative of both a direct writer and a reinsurer.

Attendees learn:• The factors that have led to this increased level of arbitration• What can be done or should have been done to avoid conflicts


Pool Reinsurance6/15, 8:30 am – 10:00 am (Open Forum)Pooling of risk is Actuarial Science 101. Insurance and reinsurance companies have sought to pooltheir risks with others as a means of sharing risk as well reducing volatility. Unfortunately, partici-pating in a pool oftentimes results in surprises and exposure to liabilities never anticipated. Thissession will discuss why many of these surprises have occurred, review some actual case studies anddiscuss what companies and reinsurers are doing to reduce and/or eliminate any mistakes of thisnature going forward.


The Reinsurance Section Council members look forward to seeing other members at the upcomingmeetings. Should you have any questions about the sessions or if you have an interest in participat-ing, please feel free to contact John Nigh, the coordinator for all sessions, at (212) 309-3958 or [email protected].

In addition to being the Reinsurance Section Council’s spring meeting coordinator, John Nigh is aprincipal at Tillinghast-Towers Perrin in New York City.

Spring Meeting Reinsurance Sessions

475 N. Martingale Road, Suite 600 • Schaumburg, IL 60173


Taking a break from planning the 2004 activities of the Reinsurance Section, members of the section council enjoy the Orlando sun.

Mike Gabon, section vice-chairperson,(left) presenting Jim Dallas, outgoingsection chairperson, with a gift ofappreciation for a job well done

Left to right: Jim Dallas (outgoing sectionchairperson), Mike Gabon, John Nigh, LeighHarrington, Tim Tongson, Ronnie Klein,Richard Lau, Larry Warren, Jay Biehl, HankRamsey, Dean Abbott (newsletter editor)

Missing: Mel Young (incoming sectionchairperson)

Author Credentials,

FSA, MAAA, is

If you are someone with good vision and enjoy the Internet, this may be the opportunity onwhich to set your sights.

The Reinsurance Section is looking to improve our section's Web site so that it provides a greater benefitto our members. To do this, we are looking for someone to become the Reinsurance Section Web Liaison.The Web Liaison will be responsible for content only—you do not have to be proficient in Web develop-ment! The Society of Actuaries' Web Department takes care of the design, the development of the site,and any other technical aspects. As Web Liason, you will interact with the Reinsurance Section Counciland the SOA staff to explore how we can best use our Web site to provide a greater benefit to themembership and then oversee the implementation of the ideas.

If you are interested, please respond to Dean Abbott at [email protected] or 612-342-3815.

reinsurance section, issue no. 52, august 2003 reinsurance news · 2012-01-19 · 2 • reinsurance...

Documents