job no.

1987 casualty actuarial society CAS journal 1987D03 [1] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES

SHOLOM FELDBLUM

Abstract

Underwriting cycles, with their wide and puzzling

swings in premiums and profitability, challenge the pric-
ing actuary to adapt rates to market realities. Under-
standing the forces behind insurance price fluctuations
is a prerequisite to analyzing market prices.
Underwriting cycles have been ascribed to actuarial
ratemaking procedures, to underwriting philosophy, and
to interest rate volatility. These interpretations underes-
timate the dynamics of the insurance marketplace, and
they ignore the competitive pressures that drive insur-
ance pricing.
Underwriting cycles, like profit fluctuations in other
industries, reflect the interdependence of rival firms.
Strong policyholder loyalty and demand inelasticity
hold the allure of large returns for incumbent firms,
but the apparent ease of entry into insurance, the lack
of market concentration, and the difficulty of mon-
itoring competitors’ prices preclude excessive prof-
its. The interaction of these forces keeps the mar-
ket in disequilibrium, with continuing price oscilla-
tions.
With the decline of rating bureaus and the growing
competitiveness of the insurance marketplace, the pro-
ficient actuary may no longer set rates based solely
on indicated costs. Insurers seek actuaries who under-
stand the competitive forces that drive market prices and
who can set future rates that are most advantageous
for the firm. They seek actuaries who can price their
products through the vicissitudes of the underwriting cy-
cle.

175
 job no. 1987 casualty actuarial society CAS journal 1987D03 [2] 09-12-02 2:43 pm

176 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

ACKNOWLEDGEMENT
The author is indebted to Benjamin Lefkowitz, Jay Siegel, and
Richard Homonoff, who suggested numerous corrections to earlier
drafts of this manuscript. The remaining errors, of course, should
be attributed to the author alone.

1. THE EDUCATION OF AN ACTUARY

When I began work as a pricing actuary, I was struck by the
simplicity of our ratemaking procedures. Actuarial techniques
are cost-based: premiums are based on anticipated losses and ex-
penses. Marketplace pricing, however, considers supply/demand
interactions, consumer desires, and competitive pressures. When
I asked about this, I was told that actuaries determine the
“proper” rates—those which best serve insurance companies and
the public.
As the months passed, I learned that insurers do not actually
set prices based on actuarial indications. Schedule rating modi-
fications of as much as 50% are used in the Commercial Lines,
and discretionary rate deviations from actuarial indications are
used in the Personal Lines. So I wondered: what is the use of
our ratemaking procedures?
When I asked about this, I was told that the poor, misguided
folk in Underwriting and Marketing always wanted lower rates.
Management was forced to cut prices below adequate levels to
keep everyone happy. Rate deviations and modifications were
the random effects of strong officers in the field.
Years later, I understood that these deviations are not entirely
random. Underwriting cycles billow through our industry, raising
and lowering the premium rates charged by insurers. The price
fluctuations are not discretionary: insurers that have ignored the
phases of the cycle have lost both money and market share. Most
important, these are industry wide cycles, unrelated to the inter-
nal politics of individual firms. Actuaries indicate rates, but the
market sets prices.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [3] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 177

Causes of the Cycle

Some actuaries believe that rates should be based only on
anticipated costs. Stable actuarial rates ensure adequate returns
for insurers, and they mitigate the price variations that anger
consumers. Carriers may be tempted by the marketing benefits
of rate cutting, but actuaries should not encourage such follies.
However, cost-based pricing is rarely optimal. Careful consid-
eration of the marketplace and of competitors’ actions is essential
for ensuring profitable operations. This aforementioned view is
dangerous to the actuarial profession as well, for if actuaries ig-
nore market realities, their companies will relegate them to tech-
nical busy-work. If actuaries wish to influence actual prices, they
must address real business concerns.
The view described and deprecated above is ensconced in two
prevalent convictions. First, underwriting cycles are seen as ex-
ternal to insurer strategies. For example, the severe downturn
in Commercial Lines operating income during the early 1980s is
sometimes attributed to high and fluctuating interest rates that en-
couraged “cash flow” underwriting. How can we price for these
variations if we can not control them or even predict them?1
Second, underwriting cycles seem unrelated to profit cycles
in other industries. Some say that insurance profits are counter-
cyclical to general business conditions: rates are high during de-
pressions and decline during prosperous periods. Others add that
underwriting cycles vary with supply restraints, not demand pres-
sures. Pricing techniques used in other industries are therefore
inapplicable to insurance ratemaking.
To understand the relationship of insurance insolvencies to
underwriting cycles, we must uncover the causes of the cycle.
Four interpretations of the cycle are described in the next section,

1 Compare Taylor [105, pg. 1]: “Individual operators in the insurance market view [the
underwriting cycle] as a variable exogenous to the formation of their own plans, one
whose timing and magnitude is beyond their control. This engenders a passive attitude
to underwriting cycles on the part of insurers.”
 job no. 1987 casualty actuarial society CAS journal 1987D03 [4] 09-12-02 2:43 pm

178 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

emphasizing (i) actuarial ratemaking, (ii) underwriting philoso-

phy, (iii) interest rate movements, and (iv) competitive strategy.
The first three imply irrational business behavior by actuaries,
underwriters, or investment officers. In addition, the actuarial
and underwriting interpretations do not explain the synchronized
pricing of independent insurers, and the interest rate interpreta-
tion cannot account for the recurrence of cycles in more stable
interest rate environments. The fourth interpretation views un-
derwriting cycles as rational business behavior among competing
firms striving to optimize long-term profits. Competition may be
rough, and it may be inexact, but it tells us a rational story if we
pause to listen.

2. INTERPRETATIONS OF THE CYCLE

Actuarial Ratemaking: Uncertainty and Counter-Cyclicality

Some actuaries ascribe profit cycles to the uncertainty and
counter-cyclicality of loss costs:

! Property/Casualty insurance costs depend upon random loss

occurrences and uncertain macroeconomic and social trends.
Random losses may be unusual weather disturbances, such
as windstorms, and earthquakes. Social trends may be unex-
pected legal changes, such as retroactive liability for pollution
exposures.
! The counter-cyclicality of insurance loss costs stems from the
time lag between the compilation of historical experience and
the implementation of new rates. Generally, two or more years
of experience are used for ratemaking, losses are developed
three months beyond the end of the experience period, systems
processing of the historical data requires another month or two,
rate analysis and filing take six months, and the rates remain
in effect for one year. Rating bureaus require an additional
half year for editing and verification of insurance data and
for notification to member companies of intended rate filings.
job no. 1987 casualty actuarial society CAS journal 1987D03 [5] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 179

Thus, the time between the average loss date in the experience
period and the midpoint of the effective period of the new rates
often exceeds three or four years (Cummins and Nye [35, pp.
232–236]).
The uncertainty and counter-cyclicality of insurance loss costs
contribute to underwriting cycles. During recessions, inflation
is moderate, automobile travel is low, jury awards are less lib-
eral, factories operate below capacity, industrial injuries are in-
frequent, and so forth.2 The experience from this period, and
the time lag between data compilation and rate implementation,
ensures moderate rate revisions for several years.
The economy soon recovers, and loss costs rise rapidly. Insur-
ers, wary of increasing their rates and losing business volume,
ascribe the mounting costs to random loss occurrences. Even
when the rate inadequacy is recognized, and rate revisions are
requested, the time lag between data compilation and rate imple-
mentation means that the needed premiums are not earned until
years later.
Historical experience continues to indicate a rate inadequacy
when the economy once again slides into a recession. Insurers

2 There are opposing influences as well. During recessions, thefts increase, leading to
higher automobile comprehensive claims. Employees recently laid off are more likely to
file Workers Compensation claims for minor injuries, since there is no loss of regular
income while on disability. Workers Compensation claim severity also increases, since
it is more difficult to find replacement jobs for injured employees (Mowbray and Black
[78, pg. 425]; Greene and Roeber [51, pp. 254–255]). For a discerning discussion of
the relationship between economic conditions and insurance loss costs in a depressed
economy, see Tarbell [104]. For relationships by line of business, see ISO [57, pg. 2],
for Personal Auto, Homeowners, and Workers Compensation, and Victor and Fleischman
[113] and Victor [112] for Workers Compensation.
Unfortunately, little is known about the correlations between insurance loss costs and
macro-economic conditions. Kahane [60], Hill [55], and Fairley [45] find that insurance
losses have a slight negative correlation with stock returns. Since stock returns reflect
economic conditions, this suggests that loss costs may be related to the economy as well.
Others find no significant correlation between underwriting returns and stock prices
(Cummins and Harrington [30]; D’Arcy and Garven [39]; Kozik [63]).
In general, the relationships noted in the text are based on conjecture and intuition.
This explanation of underwriting cycles fails for other reasons, and the absence of facts
among adherents of this theory is simply an additional flaw.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [6] 09-12-02 2:43 pm

180 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

continue filing for rate increases, even though rates have returned
to adequate levels. And so the cycle goes on.3

Awareness and Action

There are some factual problems with this interpretation. Un-
derwriting cycles are generally not counter-cyclical to macroe-
conomic conditions. Further, loss cost trends are not always dif-
ferent in prosperous times and recessionary times. But there are
more fundamental reasons why this explanation fails.
First, this interpretation presumes that pricing actuaries are
unable to learn from past mistakes and are incapable of fore-
casting loss cost trends despite years of experience. This is not
true: actuaries are proficient at estimating insurance costs and
are not easily fooled by macroeconomic conditions or long-term
social trends. Both actuaries and insurers are frequently aware
of the true loss cost trends even as rates move in the opposite
direction. For example, insurers knew that General Liability loss
costs were rising rapidly in the early 1980s, but they continued
cutting rates well below marginal cost.

Indications and Prices

Second, underwriting cycles are not due to actuarial rate in-
dications. They are due to insurer reluctance to adopt actuari-
ally recommended rate increases, to rate deviations below bureau
rates, to schedule rating credits for commercial risks, and to sim-
ilar “discretionary” rate reductions.4 Underwriting cycles are as

3 The Virginia Bureau of Insurance [114] interprets underwriting cycles in this fashion.
“The insurance cycle is usually out-of-phase with the rest of the economy. When prices
for general goods and services are rising, insurance rates are often stable and insurance
industry profits are decreasing. By the time that the rate of increase in the price for other
goods and services diminishes, data is becoming available showing that insurance rates
have not kept up with underlying costs. Insurance rates then increase rapidly and profits
improve. This lag between price increases in the insurance industry and the rest of the
economy is in large part due to the time required for claims to be reported and settled
and for claims data to be collected and evaluated.”
4 Cummins, Harrington, and Klein [32, pp. 59–60; Figure 5, pg. 59] note that “deviations
below ISO advisory rates increased substantially from 1981 through the end of 1983, as
the market softened” (see also Cummins, Harrington, and Klein [31, pg. 18]).
 job no. 1987 casualty actuarial society CAS journal 1987D03 [7] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 181

manifest in the disparity between actuarially indicated rates and

marketplace prices as in the reported net income of insurers.5
The disparity between insurer knowledge and insurer pricing
actions was particularly stark in the late 1980s, when 25% of
the Workers Compensation was being written by the involuntary
pools. Insurers were pricing the policies below cost, but they
would not write the business that they were pricing.
If disinterested analysts, uninvolved in the economic fortunes
of particular insurers, were to generate “actuarially indicated
rates” to which the entire industry adhered, there might be no
underwriting cycles. Ratemaking procedures have little or no
influence on actual profit cycles. However, insurance premium
rates are different from actuarial indications. Real-world prices
are not the result of mathematical exercises, whether simple or
sophisticated. And it is in the prices charged on the street that
we may discern the workings of the cycle.

Underwriting Philosophy
A second interpretation of insurance underwriting cycles re-
lies on the “mass psychology” of underwriters. During profitable
years, insurers grow optimistic and compete strenuously for new
business. Since capacity is limited only by financial and psycho-
logical constraints, not by physical plant and equipment, supply

5 Venezian [111] presents a more sophisticated connection of underwriting cycles with

ratemaking techniques: “Insurers and rating bureaus often use regression of past costs,
or of loss ratios, on time as a way of estimating future rate requirements. A model of
this process suggests that the rates set by such methods would create a quasi-cyclical
pattern of underwriting profit margins : : : . Empirical data on major lines of property and
liability insurance are consistent with the hypothesis that ratemaking methods contribute
to the fluctuations of underwriting profit margins.”
Venezian suggests only that ratemaking methods contribute to the cycles, not that they
cause them. But all these “ratemaking” interpretations search for the cycle in actuarial
indications where it does not exist; they ignore competitive pricing strategies, where the
cycle is powerful.
Similarly, Pentikäinen et al. [88] use a statistical model of underwriting cycles to
examine the influences of market prices on insurance solvency. Cummins and Outreville
[34] propose a model along the same lines, though with different causal variables: (a)
data collection lags, (b) regulatory lags, (c) policy renewal lags, and (d) calendar year
financial reporting.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [8] 09-12-02 2:43 pm

182 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

expands. Demand is inelastic, so premium growth means attract-

ing business from other insurers. Severe competition in a mature
market requires insurers to lower prices to gain market share
(Stewart [101]; Bloom [16]; Berger [11]).
Profits soon decline, due to low rates and the poor quality of
some risks. Underwriters become pessimistic, curtail their accep-
tance of marginal applicants, and file for rate increases. Profits
remain low until insurers re-underwrite their business and the
new rates take effect. Eventually, the rate increases and the more
careful underwriting lead to increased profits, and the cycle starts
anew.
This interpretation of the cycle is popular, and variations
abound. Boor [17, 18] suggests numerous factors that might
strengthen or weaken cycles, such as premium-to-surplus rules,
reserve management, and the ease or difficulty of entry into and
exit from the insurance market.

Information and Coordination

Should not the supply proffered and the quantity demanded
converge on an equilibrium point, and the underwriting cycles
cease? This is a central thesis of Western economics, and rapid
convergence is evident in most industries with free markets.
Stewart [101, pg. 293] explains the absence of such conver-
gence:

The cyclical process does not end for two reasons:

lack of information and lack of coordination. Indi-
vidual insurers do not and cannot know the precise
amount of insurance to supply to reach equilibrium.
They have different operating costs and, therefore, dif-
ferent break-even points or minimum acceptable mar-
gins of profit. Their perceptions and expectations of fu-
ture profits or losses develop in different ways. In self-
interest, they do not coordinate their actions. Collu-
sion, furthermore, is illegal. Even when prior approval
 job no. 1987 casualty actuarial society CAS journal 1987D03 [9] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 183

and rating bureaus had more influence on prices, in-

surers varied supply according to their own situations.

This explanation is unusual, since the lack of strategic co-

ordination and the imperfect information should lead to stable
equilibria. If firms cannot coordinate prices and quantities, then
the price mechanism effectively equates supply and demand. The
competitive characteristics of the insurance industry that Stew-
art notes argue for a more stable equilibrium, since underwriters
can quickly adjust supply to end any disparity with the quantity
demanded.6

Uniform Psychology
The fundamental problem with this explanation is not the
“lack of cooperation” or the “lack of coordination” theses. Rather
it is the assumption of a uniform psychology among underwrit-
ers. An individual may be more or less optimistic in different
years. But how is it that ten thousand underwriters across the
United States are optimistic and pessimistic in unison?

6 Stewart also cites a “cobweb” interpretation for the continuation of underwriting cy-
cles: “Cycles that result from supply’s responding to profit expectations are described in
textbook economic theory by what is called a ‘cobweb.’ : : : In agriculture, as in property-
liability insurance, demand is steady and supply is variable, with the result that prices
tend to move with changes in supply” [101, pg. 293].
On the contrary: standard “cobweb” explanations rely on the period to period lag in
revising supply. In agriculture, supply cannot be adjusted rapidly, since it depends on the
amount seeded in previous months, not just on the marketplace price. See, for instance,
Ezekiel [44, pp. 426, 436–437]: “For a commodity where the production process occupies
a definite interval of time, the period considered may be taken as so short that the total
supply available cannot be changed within the period (as, for example, the supply of
cotton or potatoes once the year’s crop is harvested),” and “The cobweb theory can apply
exactly only to commodities which fulfill three conditions: : :: (2) where the time needed
for production requires at least one full period before production can be changed : : : .” A
six-year cycle presumes a three-year production lag. This is not the case for insurance:
supply depends only on price and can be quickly adjusted.
Similarly, Cummins, Harrington, and Klein [32, pg. 63], in describing Stewart’s thesis,
write: “A key element in this explanation is that competition in soft markets ultimately
leads to inadequate rates. Prior academic research includes little or no formal analysis of
why competition could cause prices in soft markets to fall below levels needed to cover
cost expected when policies are sold and to ensure insurer financial soundness.”
 job no. 1987 casualty actuarial society CAS journal 1987D03 [10] 09-12-02 2:43 pm

184 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

Daykin, Pentikäinen, and Pesonen [40] illuminate the mystery

of cycles. Fluctuating profits are not uncommon; even random
fluctuations may look like cycles. The mystery is that while the
profit patterns in each insurer seem inexplicable, these profit pat-
terns are correlated among most of the insurers in the market.
The enigma of underwriting cycles is not that any individual
underwriter accepts risks in one year that he or she would reject
in another. Rather, it is that profits for insurers move in tandem.7
In contradistinction to Stewart’s explanation, this phenomenon
indicates a higher level of competitive strategy than we would
otherwise suspect. Insurers, no less than other firms, are sensitive
to the prices charged by their competitors, and they adjust their
own rates accordingly.
Stewart’s thesis shows the outlines of the cycle: the stable de-
mand, the competition among insurers, the fluctuating prices, and
the relatively uniform practices among underwriters at any given
time. But the connections among these phenomena remain un-
examined. To flesh out these relationships, we must ask: “What
additional characteristics of the insurance marketplace relate to
profit cycles?” and “How do these characteristics account for the
fluctuations in underwriting income?”

Cash Flow Underwriting

A third interpretation of underwriting cycles relies on interest
rate volatility. Insurers pay losses well after they collect premi-
ums, particularly in the liability and Workers Compensation lines
of business. Premiums are invested in financial markets (stocks,
bonds, mortgages) and earn investment income until losses are
paid.
Insurance income may be divided into underwriting and in-
vestment portions. Underwriting income is the difference be-

7 Daykin, Pentikäinen, and Pesonen [40] note with regard to a set of large Finnish insurers:
“The cycle is effectively the same for each of the : :: insurers, so that we can speak about
a market cycle” (emphasis in original).
 job no. 1987 casualty actuarial society CAS journal 1987D03 [11] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 185

tween (a) premium revenues and (b) loss plus expense payments.
Investment income is the return on invested assets.

Interest Rates
Interest rates rose rapidly in the late 1970s, reflecting the in-
flationary trends in the U.S. economy. Investment income be-
came a larger portion of insurance earnings, and underwriting
income decreased. Insurers wrote policies at expected underwrit-
ing losses, since they relied on investment returns for an overall
profit.
Many insurers, accustomed to underwriting profits, viewed
the reliance on investment returns as a lack of “underwriting
discipline.” They castigated this new philosophy as “cash flow
underwriting”: writing policies at a loss simply to generate pre-
mium dollars for investment.
Cash flow underwriting is appropriate as long as interest rates
remain high.8 But by the mid-1980s, new money interest rates
had fallen. The lack of underwriting discipline continued; insur-
ers kept writing policies at underwriting losses. Investment in-
come was no longer sufficient to compensate for these losses, so
insurance operating returns declined. This was the underwriting
cycle nadir of the mid-1980s.9
This argument was popular several years ago. It has lost favor
recently, since the underwriting cycle has lost no force despite

8 Compare D’Arcy and Doherty [38, pg. 86]: “While pejoratively termed ‘cash flow un-
derwriting,’ this willingness to accept underwriting losses is not a symptom of temporary
market insanity but is a rational economic reaction to the availability of higher interest
rates.”
9 See, for instance, McGee [71, pp. 22, 25]: “Changes in interest rates are the primary
force behind the recurrent swings in the industry’s profitability.” To explain the intensity
of the 1980s cycle in the Commercial Liability lines of insurance, McGee writes: “The
combined ratio for long-duration lines of insurance should move more than the ratio for
short-duration lines over the interest rate cycle, and the mix of insurance by lines will
affect the timing and volatility of the property/casualty cycle.” He acknowledges that
“workers’ compensation lines are long-tailed, but their combined ratio does not behave
as the increased interest-sensitivity principle would suggest,” although he ascribes this
anomaly to policyholder dividends and stringent rate regulation.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [12] 09-12-02 2:43 pm

186 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

the present stability of interest rates. Nevertheless, it is still useful

to examine the problems with this interpretation.

Underwriting and Investment Income

First, the distinction between underwriting and investment in-
come is specious. Cash flows must be discounted to a common
date to appropriately match revenues and expenses. True insur-
ance income is the difference between (a) premium revenues and
(b) discounted loss plus expense payments.10 True investment in-
come is the sum of (a) the return on invested surplus funds, (b)
the difference between actual and expected returns on policy-
holder supplied funds, and perhaps (c) the difference between
expected returns and the return assumed in the discount rate.11

10 Although discounted cash flows may be used to measure income, the appropriate dis-
count rate for insurance losses is unclear. Lowe [67] suggests a “negotiated rate” that
is set by the senior management of the insurance company. Woll [116] recommends an
after-tax “risk-free” rate, such as the Treasury Bill rate. Butsic [21] derives a “risk ad-
justed” discount rate based upon historical insurance experience. Fairley [45], Hill [55],
and Myers and Cohn [80] use risk adjusted discount rates, based on extensions of the
Capital Asset Pricing Model to insurance losses. The 1986 Federal Income Tax amend-
ments use the federal midterm rate to discount losses; see Gleeson and Lenrow [50] or
Almagro and Ghezzi [4]. Others have suggested embedded yields, as the Insurance Ex-
pense Exhibit uses, or new money market rates, as AICPA [1] recommends and which
most life insurers use. The lack of agreement on the appropriate discount rate hampers
consistency among insurance companies in analyzing income.
11 Compare Woll [116] and Lowe [67]. Different means of categorizing income are pos-
sible; we do not mean to prescribe a particular method. A numerical example should
help clarify the intention. Suppose the insurer has $10 billion of funds from insurance
transactions and $4 billion of surplus. Suppose also that the expected investment income
on funds from insurance transactions was 8% per annum, the actual investment income
was 9% per annum, and the investment income on capital and surplus funds was 10%
per annum; all investment income includes unrealized capital gains and losses.
Of the investment income, $800 million (or 8% of $10 billion) would be included with
insurance income, as this is part of the expected return from the insurance operations.
The remaining 1% return on the funds from insurance transactions plus the 10% return
on capital and surplus funds would be included with investment income.
Alternatively, if the loss reserve discount rate used for internal company management
reporting is 7% per annum, only $700 million (or 7% of $10 billion) would be included
with insurance income, and the remainder would be categorized with investment income.
This procedure might be used if the risk-free interest rate were 7% per annum but the
expected investment yield of the company were 8% per annum.
Numerous variants of this procedure have been suggested by actuaries. They differ
in the details—such as in the discount rates and the bases—but they all value cash
flows as of the same time. The use of unadjusted nominal values to determine insurance
profitability simply confuses performance measures and distorts patterns of profitability.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [13] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 187

When insurance income is properly measured, it is not nec-

essarily reduced by a rise in interest rates. Higher interest rates
that are accompanied by accelerating inflation increase the nom-
inal settlement values of insurance losses even as they raise the
appropriate discount rate for loss reserves. A rise in inflation
increases both investment returns and expected loss payments.
In other words, when inflation is modest, both the discount
rate and expected losses are low. When inflation accelerates, both
the discount rate and expected losses increase. The net effect is
ambiguous.12
Asset-liability matching theory also implies a different out-
come than that suggested by “cash flow underwriting” inter-
pretations of the underwriting cycle. The average duration of
Property/Casualty insurers’ assets is longer than that of their li-
abilities. A drop in interest rates, as occurred in the mid-1980s,
causes an increase in profits, not a decrease in profits. In fact,
those insurers that bought long-term bonds at high yields in the
late 1970s and early 1980s enjoyed above average investment
returns in subsequent years.13

12 For the relationship of liability losses to market interest rates, see Butsic [22]. McGee
[71, pg. 23] is aware of the inflation sensitivity of liability losses: “Inflation also has
an impact on the relationship between the competitive price of insurance and interest
rates. If costs of settling claims are expected to rise through time, a higher premium or
investment return will be necessary to cover future costs. To the extent that rising interest
rates reflect anticipated inflation, they should not affect insurance premiums.”
McGee hypothesizes that “uncertainty about the inflation outlook” in a competitive in-
dustry depresses market prices to those of the most optimistic insurer. Widely fluctuating
interest rates lead to greater uncertainty and therefore a decline in insurer profitability.
This explanation ignores McGee’s own statement that as long as inflation and interest
are correlated, different inflationary expectations should not affect insurance premiums.
Cummins, Harrington, and Klein [32, pg. 68], note that interest rate fluctuation is not by
itself a sufficient explanation of underwriting cycles: “: : : prices in competitive insurance
markets would reflect the interest earnings on funds held between the premium payment
and loss payment dates. Thus, prices should fall when interest rates rise and rise when
interest rates fall. This is not a problem unless insurers overreact to interest rate changes
or unless serious pricing errors are common.” (These remarks assume a positive equity
duration for insurers. If liability loss payments are entirely inflation sensitive, the inverse
relationship between interest rates and insurance prices does not hold.)
13 For the effect of interest rate changes on the returns of mismatched portfolios, see Bier-
wag, Kaufman, and Toevs [14] or Redington [94]. For an analysis of asset and liability
durations of Property/Casualty insurance portfolios, see Feldblum [46] and Panning [87].
 job no. 1987 casualty actuarial society CAS journal 1987D03 [14] 09-12-02 2:43 pm

188 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

Financial Expertise
Finally, and most fundamentally, a “cash flow underwriting”
interpretation of underwriting cycles reveals a deep academic
condescension towards insurance company investment managers
and underwriters. It presumes either that investment managers
were surprised by the fall in interest rates in the mid-1980s or
that underwriters are unable to adjust rates for changes in invest-
ment income. But the investment community was not shocked by
the fall in interest rates in the 1980s. On the contrary: financial
analysts were surprised that interest rates stayed high even after
inflation subsided. Similarly, good underwriters aim at long-term
operating profits. They are not easily deceived by steady changes
in investment returns.
Interpretations of the underwriting cycle abound. The ma-
jority presume that someone is erring: ratemaking methods are
naive, underwriters are simplistic, regulation is rigid, or invest-
ment managers are deceived. Such explanations search for a
cause where it is not to be found. Insurers are no less rational
than other firms are. They exist in a highly competitive market,
where the foolish firm does not long survive.

3. COMPETITION AND PROFITS

To understand the relationship of underwriting cycles to in-

surer solvency, we must briefly step aside from insurance and
delve into economics and business theory. We ask: “What is the
relationship between competition and profits?”
We consider first the economist’s perspective, examining
competitive, monopolistic, and oligopolistic market structures.
We then analyze the insurance industry from a concrete business
viewpoint, examining policy differentiation, policyholder loyalty,
and the ease of entry into the insurance marketplace. We ask:
“Given the structural characteristics of the insurance industry,
what price-cost margin should we expect?”
 job no. 1987 casualty actuarial society CAS journal 1987D03 [15] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 189

Textbook Models: Competition and Monopoly

Undergraduate economics textbooks present two market mod-
els: pure competition and single firm monopolies. These models
are meant only to illustrate the forces that determine prices, not
to depict actual practice.
In pure competition, prices are determined by industry-wide
supply and demand. No individual firm can unilaterally affect
market prices. If a firm restricts supply, its competitors take up
the slack. If a firm raises prices, consumers purchase the product
elsewhere.
In a monopolistic industry, a single firm dominates the market.
Entry of competing firms is sufficiently restricted that the mo-
nopolist can adjust the quantities supplied and the prices charged
to maximize its profits.

Competition
What market price results from each model? Suppose that
the price in a competitive industry exceeds the marginal cost of
producing the product. Any firm could cut prices slightly, garner
a greater market share, and increase its profits.
Similarly, if the market price were below marginal cost, firms
would leave the industry and employ their capital elsewhere.
Equilibrium is achieved when price equals marginal cost.
Equilibrium means that there is no tendency for prices to
either rise or fall.14 Economists maintain that prices generally

14 Industrial economists, when considering firm behavior, speak of Nash equilibria (Nash
[84]). A Nash equilibrium obtains when no firm has an incentive to modify its produc-
tion or price strategy. If firms seek to maximize their income, this implies that no firm
can obtain greater profits by raising or lowering its price or by increasing or decreasing
the quantity that it supplies. Waterson, using a game-theoretic approach to industrial eco-
nomics, defines a Nash non-cooperative equilibrium as the “point such that each player’s
strategy maximizes his expected payoff if the strategies of the others are held fixed” [115,
pg. 41]. Friedman [48, pg. 49] uses a similar definition: “A [Nash] noncooperative equi-
librium consists of n particular strategies, one for each firm, so chosen that no single firm
could possibly have obtained higher profits if it, alone, had selected a different strategy.”
Fudenberg and Tirole [49] summarize the formal theory of Nash equilibria.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [16] 09-12-02 2:43 pm

190 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

move toward equilibria in free markets. Underwriting cycles,

however, seem a stark example of disequilibrium: prices con-
tinually fluctuate.

Monopoly
Under suitable conditions, the monopolist seeking to optimize
its income will not price its product at marginal cost.15 When
price equals marginal cost, there are no economic profits for the
firm. But if the monopolist restricts output, consumers “bid up”
the price to obtain the scarce good. Price exceeds marginal cost,
and the firm receives additional profits.
In a purely competitive marketplace, price equals marginal
revenue which equals marginal cost. In a monopolistic market,
marginal revenue generally exceeds marginal cost. Prices are
higher in a monopolistic market than they would be in a com-
petitive market.

Actual Market Structures

These market structures rarely exist in their ideal forms. Even
when there are thousands of firms selling similar products, com-
petition is seldom perfect. For instance, grocery stores exist all

When market conditions cause firms to have different strategies—some seek stable
current income and others seek to increase sales—Nash equilibria often dissolve. This
phenomenon underlies the model of underwriting cycles developed below.
15 These conditions are that either the marginal cost rises as quantity supplied increases or
the demand curve slopes downward. Marginal cost is the cost of producing an additional
unit of the good. In insurance, this is the expenses and anticipated losses of writing
an additional policy, not the average expenses and losses incurred on the current book
of business. The demand curve is the relationship between consumer demand and the
product’s price. In insurance, this is the number and size of policies and endorsements
desired by consumers at each premium rate.
Both conditions are satisfied in the insurance market. (1) The demand curve in many
lines of business is nearly vertical, because of statutes, regulations, and business policies
that mandate coverage (Sherdan [99]). (2) The marginal cost curve rises sharply, despite
the preponderance of variable costs in insurance. As D’Arcy and Doherty [38, pg. 9]
note: “: : : an insurer writing a large quantity of policies will eventually have to relax
underwriting standards to increase the quantity further, and the newer policies could have
a higher expected loss ratio.” That is, at low quantities, insurers can “skim the cream,”
selecting the best risks. At higher quantities, insurers offer coverage even to mediocre
and poor risks. Thus, marginal costs rise as the number of policies issued increases.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [17] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 191

over, selling the same foods: is this not pure competition? But
most individuals use the nearest corner grocery for small pur-
chases and do not bother to price shop. In other words, the gro-
cery store may have a near monopoly within a small neighbor-
hood.16
Monopolies are equally hard to maintain. IBM dominated the
market for mainframe business computers in the 1960s, and it
enjoyed large price-cost margins during those years. But com-
petitors soon entered wherever profits beckoned—computer pe-
ripherals, software programs—and they quickly gained signifi-
cant market shares.17
Nevertheless, these two models are important, for they set the
bounds of the price range. If capital can be transferred to other
uses, firms will not price below marginal cost.18 And if suffi-
cient supply is available, firms will not price above the monopoly
price.

16 Scherer [97, pg. 325] comments: “Even when firms produce physically identical com-
modities, complete homogeneity is not likely to be attained because of differences in
location : : : . When producers are located at different points on the map, their products
are said to be spatially differentiated.”
17 On the history of IBM’s market dominance in the mainframe computer industry and
the entry of competitors in peripheral equipment and software products, see Brock [20].
Government sponsored monopolies, such as municipal utilities, cable TV franchises, and
telephone service until the 1980s, are different. These industries have strictly regulated
rates; they do not price by supply and demand considerations.
The diversity of insurance rate regulation affords a range of insurance markets. In
some states, such as Massachusetts and Texas, insurance rates are set by the regulator
or by official rating bureaus. In other states, such as Illinois and pre-1989 California,
the free market determines insurance prices. Insurance rate regulation is a factor (albeit
a minor one) in underwriting cycle severity.
18 Transferring capital can be difficult, and firms may price below marginal cost in a
declining industry. The Personal Insurance lines present an excellent illustration of this.
Over the past 40 years, direct writers have steadily garnered most of the Personal Lines
market, and they have consistently attracted the better risks among the insured pop-
ulation. Independent agency companies have a declining market with worsening risk
quality. Many of these companies are slowly moving to other lines of business (such
as Commercial, Specialty, Reinsurance, and Substandard Auto), experimenting with less
expensive distribution systems (such as direct mail), or trying to start joint ventures with
other financial institutions (such as life insurers, health insurers, and securities brokers).
Meanwhile, average Personal Lines returns for independent agency companies are below
marginal cost.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [18] 09-12-02 2:43 pm

192 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

This price range is wide, since the monopoly price may be

well above marginal cost. So if the market is neither purely com-
petitive nor monopolistic, what prices will actually be charged?

Economic Models
Economics is rigorous. Theorists provide the needed assump-
tions, then “prove” the desired conclusions. But these assump-
tions are invariably idealistic. The equations are mathematically
perfect but of limited practical value.
We cannot proceed without a theoretical framework. We will
deal with price-cost margins, Nash equilibria, entry conditions,
and price elasticity of demand. However, we are interested not
in formulating theorems but in understanding a business phe-
nomenon: the underwriting cycle. So we must step gingerly over
the coming terrain.
We can view this distinction from another perspective. Eco-
nomic models abstract reality. They isolate some elements, and
the results are determined from the assumptions. The business
world is represented by succinct mathematical expressions.
Underwriting cycles, however, are complex phenomena: no
two companies react identically to their course. We will not try to
determine the exact duration or severity of the cycles. Rather, we
seek to understand the driving forces behind insurance pricing.
We begin with an abstract model of pricing in a competi-
tive market with a limited number of firms.19 Our emphasis will

19 In truth, there are thousands of American insurers, and dozens of new ones enter the
industry each year. This is a central characteristic of the insurance model that we develop
further on. For clarity of exposition, however, we begin with a model of a limited number
of firms.
Supplier interdependence is enhanced by high market concentration. Some economists
use four firm concentration ratios of 50% or greater, or a Herfindahl-Hirshman index of
about 1,000 or greater, as indicators of possible interdependence. (See, for example,
the June 1984 Justice Department merger guidelines for antitrust action.) The Personal
Auto insurance industry shows a four firm concentration ratio of 40% and a Herfindahl-
Hirshman index of 610 on a national basis, and corresponding average figures of 53%
and 1,000 on a statewide basis. These figures depend on the definition of the market:
 job no. 1987 casualty actuarial society CAS journal 1987D03 [19] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 193

be on Nash equilibria and conjectural variation. We then revise

the model, discarding the idealism and adding reality, to explain
profit cycles in the Property/Casualty insurance industry.

Conjectural Variation
Suppose two rival firms, producing identical products, each
have 50% of the market. Consumers are conscientious price
shoppers with excellent information, so if either firm underprices
the other it quickly captures the entire market. If the firms com-
pete by setting prices, then a static microeconomic analysis im-
plies that both firms will set prices at marginal cost.20 If one firm
prices above marginal cost, the other firm can charge slightly
less, gain the other 50% of the market, and increase its total
profits.

state versus national and individual line versus all insurance products. Inter-industry
comparisons of market concentration must use similar criteria of market definition; if so,
insurance shows low relative concentration. On automobile insurance, see Klein [61, pg.
12, Table 1, pp. 18–19, Table 4]; on Workers Compensation insurance, see Countryman
[29, pg. 17, Table 1], Klein [62], and Appel and Gerofsky [6; 7].
20 Firms may compete either by setting prices or by choosing the quantities they supply.
Price and quantity are interrelated, since the industry demand curve sets a one-to-one
relationship between them. If firms compete by choosing the quantities they supply,
“Cournot competition” implies that the resulting price will exceed marginal cost. The
price-cost margin varies inversely with the number of firms: one firm (pure monopoly)
produces the greatest profits, and an infinite number of firms (pure competition) elimi-
nates economic profits. See Tirole [107, pp. 218–221], or Scherer [97, pp. 152–155].
Manufacturing firms with long production cycles may compete by choosing the quan-
tities that they supply. A Cournot analysis is appropriate for them. Insurers have almost
no supply restrictions; rather, they compete on premium rates. A “Bertrand” analysis,
which results in price equaling marginal cost, is the appropriate model (see below in this
note). See Tirole [107, pp. 209–212] or Varian [109, pp. 461–464].
The appropriate model for insurers depends on their supply constraints. Unlimited
capacity implies that firms compete by setting prices. Severe capacity constraints imply
that firms compete by choosing quantities. For an analysis of the limits on insurance
capacity, see Stone [102]. Stone’s analysis applies to large Commercial risks, where
random losses may adversely affect an insurer’s income or even solvency. In practice,
there are no capacity constraints in the Personal Lines or for small Commercial risks.
Moreover, for some large risks, the availability of reinsurance mitigates the capacity
constraints.
For a general discussion of insurance supply, see Stewart [101]. Stewart correctly notes
that insurance supply is determined by psychological and financial considerations, not by
plant, equipment, labor, or other physical restrictions. The ability of insurers to quickly
revise quantities and prices is an essential aspect of the underwriting cycle; see the text
below.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [20] 09-12-02 2:43 pm

194 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

This analysis is static: it considers only a single time period.

Dynamic models presume that firms respond to their rivals’ com-
petitive actions. Moreover, each firm anticipates how its rivals
will respond before implementing its own strategy. Economists
term this conjectural variation: “Each firm believes that its choice
of price will affect the price selected by its rivals.”21
Suppose again that two firms producing identical products
and competing on price each have 50% of the market. In the
static analysis, if the market price exceeds marginal cost, then
either firm may slightly reduce its price and garner the entire
market. In reality, the businessman wonders: “If I cut my price
to increase market share, how will my rival respond?”
Clearly, the rival will match the price cut—at least if a small
reduction in price enables it to retain its market share. If both
firms presume that the other will match a price cut, neither will
initiate the price reduction.22
We formulate this mathematically as follows. Let P m be the
current market price and P c be the competitive, or marginal cost,
price. Let v be the annual discount rate for future earnings (the
discount rate is treated more fully below). Suppose that each
firm knows that if it reduces its price below P m , its rival will
immediately charge P c . Finally, assume that a price cut below
the current market price promptly attracts the entire consumer
population.23
The current market price, P m , provides total industry earnings
of E m , a positive amount. The marginal cost price, P c , provides

21 Tirole [107, pg. 244]. For a mathematical development, see Varian [110, pp. 102–103],
or Waterson [115, pp. 18–19]. Porter [93] presents a non-mathematical discussion of the
strategic consideration of expected rival responses.
22 That is, conjectural variation influences optimal business strategy. If an insurer believed
that its peer companies use cost-based pricing and that they do not consider competitive
pressures, it would have no disincentive to reduce rates in order to gain market share.
In practice, insurers’ prices are strongly affected by those of their rivals. This is most
evident in the Personal Automobile market, where the major direct writers carefully
examine their rivals’ rates, by territory and classification, to set their own prices.
23 These are the ideal assumptions so endearing to economists. We will return to reality
in a few paragraphs.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [21] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 195

zero economic profits; that is, E c = 0. If both firms maintain

the current market price, P m , their earnings will be 1=2(E m +
vE m + v2 E m + " " " ) for each. If either firm slightly shades prices,
its earnings will be E m in the current period.24 Since its rival
quickly cuts prices to marginal cost, its earnings are 0 in all
future periods.
If the firms are to be dissuaded from cutting prices, then E m
must be less than 1=2(E m + vE m + v2 E m + " " " ). That is,
1 < (1 + v + v2 + " " " ) # 2, or v > 12 :
This makes sense. If v is high enough (more than one half in
this instance), firms are unwilling to sacrifice future earnings for
immediate profits. Conversely, if v is low, firms disregard future
earnings and emphasize short-term results.25

Discount Rates
The discount rate measures the relative value of a dollar of
future earnings compared with a dollar of present earnings. The
interest rate is a part—but only a part—of this. Also important
is the uncertainty about future market conditions. Perhaps con-
sumer demand will slacken, other suppliers will enter the indus-
try, restrictive regulations will impede price adjustments—and
future profits will dissipate. Perhaps demand will grow and en-
try barriers will harden, increasing future profits. Perhaps rival
firms will differentiate their products and segment the market.26
Future earnings in an inflationary economy are worth less in
real dollars. In a competitive market, they are also uncertain:

24 This is a theoretical model. It assumes that an infinitesimal price reduction attracts the
entire market. In insurance, (1) a substantial rate reduction is required to gain market
share, and (2) shifts in the insured population occur at renewal time, not continuously.
The model of underwriting cycles developed below incorporates these elements.
25 For more complete discussions, see Tirole [107, pp. 245–251], or Shapiro [98].
26 Describing the discount rate, ±, Shapiro [98, pg. 362, note 58] writes: “Formally, ±
may be thought of as the product of two terms: ± = ¹e$iT , where ¹ is the hazard rate
for the competition continuing (i.e., the probability that the game continues after a given
period, given that it has not previously ended), and e$iT is the pure interest component
of the discount factor, with period length T and interest rate i.”
 job no. 1987 casualty actuarial society CAS journal 1987D03 [22] 09-12-02 2:43 pm

196 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

anticipated profits may never materialize. Business strategy,

which determines the quantities supplied and the prices charged,
affects the realization of future profits.
The size of the discount rate (v) needed to discourage price
cutting varies with the number of competing firms. If there are
two firms of equal size, v must be greater than 12 , as the equation
above implies. If there are ten rival firms of equal size, v must
be greater than 9/10 to discourage price cutting.27 The insurance
market has hundreds of rival firms in the major lines of business,
so this simple model implies that the discount rate must be near
unity to discourage price cutting. But if insurers generally price at
marginal cost, why are there severe profit cycles? To answer this
problem, we present a more sophisticated model. First, however,
let us take another detour: How does a firm choose an “optimal”
price?

Limit Pricing and Entry Barriers

The optimal price depends upon the strength of entry barriers.
If entry barriers are low and profits are high, new firms enter
the market. Entrants cannot gain market share if they charge
the current price, so they have little to lose by price cutting.28
Incumbent firms rarely let the market price remain high enough
to attract new entrants.
The cut-off price between attracting and discouraging new en-
trants is termed the “limit price.” But why should the limit price
be any different from the competitive marginal cost price? If all
firms have the same production costs, then any price exceeding
marginal cost attracts new entrants.29

27 That is, E m must be less than (E m + vE m + v 2 E m + "" " )=10. Thus, 1 < (1 + v + v2 +
" " ")=10, or v > 9=10.
28 In underwriting parlance, we speak of new entrants “buying” market share. A new
firm may suffer operating losses for several years before it develops a profitable book
of business. This is particularly true in insurance, since new entrants attract the marginal
and unprofitable risks.
29 “Limit pricing” is a standard economic term, unrelated to the actuarial procedure of
“increased limits pricing.”
 job no. 1987 casualty actuarial society CAS journal 1987D03 [23] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 197

But firms do not all have the same production costs. In par-
ticular, new firms face a fixed (sunk) cost of entry, so the limit
price exceeds the marginal cost price.30
In theory, there are few barriers to entry in insurance. The
insurer need build no factories to manufacture its product; it
may contract for the needed actuarial, underwriting, and loss ad-
justment skills; and statutory capitalization requirements are not
excessively onerous (although they are higher than they were
before the advent of risk-based capital requirements). The firm
may simply “hang out a shingle” and begin writing policies.
In practice, this is not correct. In the Personal Lines market,
the direct writers are profitable whereas the independent agency
companies are losing money. Yet few independent agency com-
panies have successfully switched to direct writing or exclusive
agency distribution systems. The constraints on the distribution
system are powerful, raising large entry barriers to the profitable
insurance markets.31
The traditional barriers to entry, such as minimum efficient
production scales, or the advertising budget needed to place
products on retail shelves, are not important in insurance. The
insurance “distribution” barrier to entry does not involve getting
consumers to purchase policies. Rather, it involves getting the
better risks to purchase policies.
We return to this topic later on, in our model of underwriting
cycles. Note, however, how deceptive these barriers to entry are.

30 On limit pricing, see Milgrom and Roberts [76] and Porter [93, pg. 14] (who uses
the term “entry deterring price”). Insurers face few fixed costs, particularly in lines of
business dominated by the independent agency distribution system. Entry into the Com-
mercial Lines insurance marketplace is deceptively easy—new firms believe they can
enter quickly. Thus, there is a short span between the marginal cost price and the limit
price.
31 Several life insurers have recently entered the Property/Casualty Personal Lines mar-
ket. Although they came with strong underwriting, actuarial, and distribution systems,
enormous capital, and extensive experience in Life and Health insurance, most of these
firms have had trouble transforming the newly acquired Personal Lines risks into prof-
itable books of business. The hidden barriers to entry are strong deterrents to prospective
insurers.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [24] 09-12-02 2:43 pm

198 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

It is easy to enter the insurance market, since there are no ma-

jor capital or regulatory barriers. It is far more difficult to enter
successfully.
These are the bounds postulated by industrial economics. In
the long run, prices will not remain below marginal cost or above
the limit price.32 The actual prices charged depend on the number
of firms, the extent of “conjectural variation,” the discount rate
assumed by each firm, and other factors affecting the price-cost
margin.
The theoretical economist would ascribe the insurance indus-
try’s low profitability to the competitive characteristics of its
market.33 But we need a more specific analysis to understand
underwriting cycles, so we ask: “How do the nature of the insur-
ance product and the operations of the insurance carrier affect
anticipated profits?”

4. INSURANCE INDUSTRY CHARACTERISTICS

An industry’s structure and the characteristics of its products

influence both expected profits and strategic possibilities. Three
considerations particularly germane to insurance are

1. Product differentiation and substitute products,

2. Cost structures and barriers to entry, and
3. Consumer loyalty and price shopping.

We begin with these insurance attributes, in preparation for the

analysis of underwriting cycles.

32 In the short run, this is not true. In declining industries, prices often sink below marginal
cost. In expanding industries, incumbent firms may price above the limit price, allowing
new entrants even as they reap large profits. Numerous other short term exceptions are
discussed in the economics literature.
33 Plotkin [89, 90, 91, 92] has documented the relative profitability of insurers vs. other
firms. See also Braithwaite [19], Banfield [9], and Bailey [10].
 job no. 1987 casualty actuarial society CAS journal 1987D03 [25] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 199

Product Differentiation
When firms supply products with varying attributes, such as
automobiles, computer hardware, and fashion clothing, each of
them may enjoy some market power and associated economic
profits. When the products of competing firms do not vary much,
as is true in agriculture, all firms may be constrained by the prices
of the most efficient producer. In short, product differentiation
increases expected profits.
Most insurance policies are indistinguishable to the average
consumer. In certain lines of business, such as Workers Com-
pensation and no-fault Auto Insurance, benefits are mandated by
statute. Even where no laws impede differentiation, product di-
versity is hard to maintain. Improved policy forms can be copied
by rivals, so advantageous innovations are transient.
The existence of close substitutes for an industry’s products
has a similar effect: substitutability constrains profitability. For
instance, aluminum often can be substituted for steel. Aluminum
prices constrain steel profitability, regardless of competition in
the steel industry.
In many lines of business, there are few substitutes for insur-
ance. The Personal Lines consumer has no choice but to pur-
chase an auto insurance or Homeowners policy. Similarly, most
small business owners must buy Workers Compensation insur-
ance, since self insurance techniques are feasible mostly for large
and sophisticated companies. The rising claims consciousness of
the public, and the increasing predilection of Americans to turn
to the courts, strengthens the demand for Commercial Liability
products. Small businesses have no alternative other than to buy
insurance protection.
In sum, the lack of product differentiation means that indi-
vidual insurers have difficulty increasing prices and profits. But
the lack of close substitutes for an essential product means that
the industry as a whole can raise or lower premium rates without
losing consumer demand. Formally, aggregate consumer demand
 job no. 1987 casualty actuarial society CAS journal 1987D03 [26] 09-12-02 2:43 pm

200 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

for insurance products is inelastic with respect to price, but inter-

firm elasticity is high.

Cost Structures and Barriers to Entry

We distinguished above between traditional and “hidden” bar-
riers to entry. Traditional barriers depend on cost structures: min-
imum efficient plant size, up-front capital requirements, the time
needed to enter, and production process learning curves. Poten-
tial entrants observe these costs, which influence their willing-
ness to join the industry.
Insurance has few traditional barriers to entry. Almost all
costs, including losses, loss adjustment expenses, commissions,
salaries, and premium taxes, are variable, not fixed.34 No plants
need to be built, no expensive equipment is required, and statu-
tory capitalization requirements are manageable.35 Most costs
are paid either on the policy effective date (e.g., commissions)
or after the policy is in force (e.g., losses).36 The cash inflows
from “producing” an insurance policy precede the cash outflows,

34 The distinction between variable and fixed costs differs from the actuarial distinction
between costs that vary directly with premium and those that do not. Salaries of non-
managerial personnel are variable costs, though they do not vary directly with premium.
The other expenditures listed in the text are both variable costs and vary directly with
premium.
35 Meyerson [74, pg. 151], writing before the advent of risk-based capital requirements,
notes that “the initial capital and surplus requirements of most states are much too low un-
der present conditions.” Danzon [36] examines the relationship of state licensing statutes
to entry barriers, in terms of delay of operations and cost of entry. She finds average
delays of six to ten months, and an average personnel cost per state for entry expenses
of $100,000. She notes that these costs are too small to serve as entry barriers. See also
Klein [62, pp. 91–92], who shows high entry and exit to the Workers Compensation
market.
The implementation of risk-based capital requirements in 1994 for Property/Casualty
insurance companies should somewhat raise these entry barriers. For some small insurers,
though, the risk-based capital requirements are not that much higher than the previous
minimum capital requirements. The effect of the new capital standards is more evident
for medium and large insurers. In fact, an early attempt to add a “small company charge”
to the risk-based capital formula died on the conference table in 1993.
36 Other acquisition expenses and certain administrative and underwriting costs are ex-
pended before premiums are received. The National Council on Compensation Insurance,
using a 1977 study of Massachusetts Workers Compensation expenses, estimates that only
14% of “other expenses” (i.e., general expenses, other acquisition costs, and miscella-
neous taxes, licenses, and fees; thus, about 2% of insurance costs) are paid before the
 job no. 1987 casualty actuarial society CAS journal 1987D03 [27] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 201

thereby facilitating the entry of new firms. Underwriting intrica-

cies are not readily discernable, and many entrants believe that
there is no significant learning curve. (In fact, casualty under-
writing is a fine art, but new entrants sometimes seem loath to
admit this.) Finally, a firm can contract for underwriting, actuar-
ial, accounting, and loss adjustment skills, so little time is needed
before writing policies.
As we noted earlier, the “hidden” barriers to entry in insur-
ance are powerful. It is easy to enter the insurance marketplace;
it is far more difficult to enter successfully. New entrants attract
marginal risks, and actual insurance losses are high in early pol-
icy periods. It takes many years to obtain a profitable book of
business (Conning & Co. [27]).
So new firms continuously enter the insurance market. Were
earnings steady, the high rate of entry would depress expected
profits. But fluctuating earnings, and the “hidden” entry barriers
discussed above, impair the chances of successful operations.
Many new entrants, with low quality books of business, do not
last through the trough of the first underwriting cycle.

Consumer Loyalty
Price changes affect purchasing decisions. If the price for a
particular brand of toothpaste rises 10%, some buyers of that
toothpaste may switch to other brands.
Some goods have large “switching costs.” Consumers of large
electrical equipment may not change suppliers unless prices rise
substantially, since such a switch would involve costs of installa-
tion, inspection, testing, retraining, and adapting other machin-
ery. In other words, consumer loyalty to a particular brand or

policy’s inception; see WCRIBM [117]. Mahler [68, Appendix 11, pp. 269–270] esti-
mates that only 20% of “company expenses” (that is, general expenses, other acquisition
expenses, and one half of unallocated claim expenses; thus, about 3% of insurance costs)
are paid prior to the policy’s inception.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [28] 09-12-02 2:43 pm

202 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

supplier depends upon the costs of changing products.37 High

switching costs impede competition and raise expected profits.
Toothpaste, unlike large electrical equipment, has no “switch-
ing costs.” Consumers have no constraints, either ante hoc or
post hoc, on the brands they choose. When switching costs are
absent, competition more easily dissipates economic profits.
Insurance seems similar. At renewal time, a consumer can
purchase coverage from a competing carrier with no additional
costs or gaps in coverage. This implies low expected profits in
insurance.
In truth, insurance is not at all like toothpaste, particularly in
the Personal Lines. Insureds rarely compare competitors’ prices
when their policies come up for renewal, whether or not they
made such comparisons when they first obtained the coverage.38
Only if an insurer dramatically raises its rates will policyholders
begin searching for other agents or carriers.
Over the long term, insurance is no different from other goods.
Higher than average prices cause a slow but steady loss of mar-
ket share, which is extremely difficult to win back. But in the
short term, a reputable insurer can maintain a higher than average
price-cost margin without a significant loss of business.
Were insurance earnings steady, long-term expected profits
would be low. The lack of product differentiation and the appar-
ent ease of entry would force insurers to price close to marginal
cost. But the lack of close substitutes, consumer loyalty, and the
difficulty of successful entry facilitate short-term price fluctua-

37 Porter [93, pg. 10] defines switching costs as “one-time costs facing the buyer of
switching from one supplier’s product to another’s”; he adds: “Switching costs may
include employee retraining costs, cost of new ancillary equipment, cost and time in
testing or qualifying a new source, need for technical help as a result of reliance on seller
engineering aid, product redesign, or even psychic costs of severing a relationship.”
38 Fox [47] reports that most of the auto policyholders who made cost comparisons did so
at least two years prior to the survey date; see particularly his Tables 2 and 3 on page 23.
Joskow [58] describes the relationship of policyholder information to insurance industry
market structure.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [29] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 203

tions. These characteristics of the insurance industry underlie the

model of underwriting cycles in the following sections.

5. DYNAMICS OF THE UNDERWRITING CYCLE

Profit Cycles
Profit fluctuations may take two forms. In the first form, the
market is in equilibrium during certain periods. External influ-
ences change costs, supply, or demand, and they thereby shift ex-
pected profits. Disequilibrium results until the price mechanism
forces profits back to the original level. If external influences
again affect the market, the fluctuations start anew.
Such profit fluctuations are rarely cyclical. For instance,
weather conditions affect farm produce and profits: an unex-
pected frost may damage citrus fruit production, or a severe
drought may lower crop supply. The affected farmers suffer from
lost production, while other farmers benefit from higher prices.
Prices and profits fluctuate, but the pattern is not cyclical.
Underwriting cycles take a different form: no phase is in
equilibrium. Insurer strategies during profitable years drive rates
down; changed strategies during poor years push rates up.
At two points in the cycle, in the upswing and the downturn,
prices pass through the same point. But the underlying forces are
different. One reflects a downward driving price strategy founded
on high rates; the other reflects an upward driving price strategy
founded on inadequate rates. This difference may be missed by
an outsider looking at a snapshot of industry income. But the
disparity is keenly felt by the businessman struggling for profits.

The Profitable Years

If there is no equilibrium point, there is no good place to
begin analyzing the cycle. Yet we must start somewhere. So we
begin, perhaps arbitrarily, at the top, as in 1977–78 or 1986–87
 job no. 1987 casualty actuarial society CAS journal 1987D03 [30] 09-12-02 2:43 pm

204 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

or 1992–94 (for Workers Compensation): income is high and

insurers are satisfied.

Entry and Exit

Satisfaction breeds desire. Outside firms are enchanted by the
ease of insurance operations: simply write the policy, collect the
premium, and pay less in claims while you invest the assets.
There are few explicit barriers to entry, so new firms join the
industry.
Figure 1 shows insurance company entries and exits in the
1980s. Note the prevalence of entry into an industry earning
below average profits and with low growth potential. Many of
these entrants quickly failed. Insurance company exits climbed
during the unprofitable 1984–85 and 1989 periods, and dipped
in the profitable 1980–82 and 1987 periods.39
New insurers cannot sell their policies at the going market
rate. Entrants must discount prices in any industry. This is all the
more true in insurance, where it is hard to attract new customers.
But new insurers believe that they have little to lose by charging
lower rates. They have no existing business, so they do not lose
money on older policyholders by cutting rates. Any price above
marginal cost is profit.40

Price Shaving and Market Shares

New entrants charging low rates are an unwelcome thorn in
the industry’s side. Equally unwelcome is the change in strategy
among existing insurers.
The model presented in Section 3, “Competition and Prof-
its,” assumes an equal division of the market among insurers.

39 See Stern [100]. Nelson [85], analyzing data for 1957 through 1967, notes that the
number of exits is correlated with the combined ratio with a lag of one year.
40 Meidan [73, pg. 395], who calls this a “market challenger strategy,” notes that it “is
characterized by the aggressiveness of the marketing tactics. Typically insurers that follow
this strategy are ambitiously trying to grow as fast as they can.”
job no. 1987 casualty actuarial society CAS journal 1987D03 [31] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 205

 job no. 1987 casualty actuarial society CAS journal 1987D03 [32] 09-12-02 2:43 pm

206 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

Suppose, instead, that there are ten firms: one has 50% of the
market, eight have 6% of the market, and one has 2% of the
market. Also assume that the appropriate discount rate is 10%
per annum. Let us restore the ideal assumptions for a moment:
if any firm cuts prices, it immediately attracts all consumers.
Moreover, if any firm cuts prices, its competitors reduce their
prices to marginal cost.
The large firm presently earns 50% of the industry’s economic
profits. If current pricing continues, it will earn this amount in
perpetuity. Using the notation of Section 3, where E m is annual
economic profits at the market price and v is the discount rate, the
present value of this profit stream is (50%)(E m )(1 + v + v2 + " " " ).
This equals 5:5 % E m at a discount rate of 10%. If the insurer cuts
prices slightly, it earns a bit below E m in the current year, but
no economic profits in all future years. The large firm has an
incentive to continue its present pricing strategy.
Now consider the firm with only 2% of the market. It now
earns 2% of the industry’s economic profits. If conditions do not
change, it will earn this amount in perpetuity. The present value
of its profit stream is (2%)(E m )(1 + v + v2 + " " " ), or 0:22 % E m at
a 10% discount rate. If it cuts prices slightly, it earns much more
than this in the current year. The small but aggressive firm has
a strong incentive to cut prices.41
Realistically, of course, the small insurer will not instantly
capture the entire market with a small price reduction. Most pol-
icyholders are loyal to their current insurers, and they often ig-

41 Harrington and Danzon [54] suggest that the aggressive marketing strategy of small
firms may result from an inability to avoid the “winner’s curse.” In competitive bidding
among suppliers, a firm which provides unbiased bids will generally win only when its
offered price is too low. When its offered price is too high, another supplier will generally
win. Harrington and Danzon differentiate between established and inexperienced firms:
“: : : established firms in stable markets have learned to make formal or informal adjust-
ments to their loss forecasts in order to avoid the curse. The availability of information
from agents and brokers also may facilitate this process : : : . Inexperienced firms may use
nonoptimal forecasts, placing too much emphasis on their own information or drawing
incorrect inferences from the actions of other firms.”
 job no. 1987 casualty actuarial society CAS journal 1987D03 [33] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 207

nore competitors’ rates at renewal time. The small firm’s rate

decrease would slowly increase its market share: say, 10% a year.
Although substantial, the gain is not overwhelming.
The large insurer expects different outcomes. A carrier with
50% of the market may have already saturated its target customer
populations. Even if it desires to grow rapidly, there are few new
insureds for it to attract. The large firm’s rate reduction may
increase its market share only 1% a year.

Rival Responses
Competitive responses to rate cuts by a small firm or a large
firm also differ, particularly in insurance. Premium rates vary by
classification, territory, type of coverage, and similar dimensions.
Rate comparisons can be an exhausting task, especially when
the classification schemes of the insurers differ. Thus, carriers
do not monitor premium rates of small companies. In Personal
Auto insurance, insurers analyze the rates charged by State Farm,
Allstate, and a handful of other large carriers. The premiums
charged by smaller insurers are revealed only in industry-wide
accounting statistics. Actual rates, although publicly available in
rate filings, are rarely examined.
Moreover, rivals do not react swiftly to rate cuts by small
insurers. If a firm with 1% of the market has a 10% growth in
business, and the new business is drawn evenly from its rivals,
then the other firms suffer only a 0.1% decrease in volume. If an
insurer with 50% of the market has the same growth, its rivals
lose 10% of their business.
Thus, when rates are high, small insurers are tempted to
cut prices aggressively.42 Their actions may not be noticed, re-

42 Anderson and Formisano [5], in a study of six insurance failures between 1975 and
1985, found rapid premium growth, expansion to other states, and inadequate pricing to
be three of the most significant causes of the insolvencies. For instance, in the years pre-
ceding the insolvencies, Reliable Insurance Company and All-Star Insurance Company
had premium growth of over 50% per annum. Wisconsin Surety Company expanded
 job no. 1987 casualty actuarial society CAS journal 1987D03 [34] 09-12-02 2:43 pm

208 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

sponses of rivals will be delayed, they may increase market share

rapidly, and their revenues will climb. Large insurers, however,
have less incentive to reduce rates. Their market shares increase
more slowly, their actions are quickly noticed, competitors re-
spond swiftly, and the premium lost on existing business may
exceed the premium gained on new insureds.
The incentive for an incumbent insurer to reduce rates de-
pends on the expected profits in its renewal book of business.
Renewal business is generally more profitable than new busi-
ness, and insurers strive to maintain policyholder loyalty. An
incumbent insurer may reduce its own rates to avoid the loss of
profitable renewal business to a competitor.
The profitable phase of the underwriting cycle is in disequi-
librium. Some firms enjoy current earnings, others aggressively
seek to grow, and entrants clamor to join the industry.

Competitive Strategies
Profits influence business strategies. As the profitable phase
of the underwriting cycle continues, more firms ignore short term
income and seek growth. For simplicity, let us differentiate strate-
gies between (a) aggressive growth and (b) price maintenance.
Assume that at time t, w% of firms emphasize aggressive growth
and (100 $ w)% of firms emphasize price maintenance.
The change in w depends upon the sign and magnitude of
economic profits, labeled p here. The greater the economic prof-
its and the longer the economic profits are expected to persist,
the more firms will seek aggressive growth.43

from 2 states to 13 states in 6 years, and Eastern Indemnity Corp. expanded from 1 state
to 34 states in 5 years. The aggressive marketing strategies of these insurers eventually led
to their failures. As Anderson and Formisano comment (page 460): “rapid growth : : : can
realistically only be accomplished by pricing below cost and taking an unreasonable
proportion of poor risks.” Similarly, Best’s [12, pg. 39] notes that “approximately 81%
of all insolvencies occurred in companies experiencing unusual growth trends, which we
defined as growth outside industry norms of 5% to 25%.”
43 Actuaries are tempted to express such relationships as partial derivative equations. We
might say that the partial derivatives of w with respect to both t and p are positive.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [35] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 209

This price maintenance strategy is not sustainable. If your ri-

vals are cutting prices and gaining market share, you must either
respond or disappear. But the optimal response depends on the
number of firms reducing rates. If the percentage of firms ag-
gressively seeking market share is small, then it is reasonable to
hold prices above marginal cost. The high level of policyholder
loyalty to the insurer means that insurance market share growth
is a slow process. For instance, suppose that 10% of firms are
aggressively cutting rates, or w = 10%. (For simplicity, assume
that firms are of equal size, so 10% of firms means 10% of the
market.) If such discounts provide a 10% annual growth in mar-
ket share, then these firms will have 11% of the market after a
year’s time, and their rivals will remain with 89% of the market.
The maintenance of high prices has led to a 1% reduction in
market share—a small loss compared to current profits.
If 50% of firms are aggressively reducing prices, the outcome
changes. The same 10% market share growth for these firms
reduces their rivals’ portion from 50% to 45%. Short term profits
do not offset a 10% loss of business.

The Nadir of the Cycle

How might one respond? Following rates downward is no
remedy. The insurance industry has thousands of firms, a com-
petitive structure, and invitingly easy entry conditions. Expected
profits would be extremely low if prices were left purely to mar-
ket pressures.
Indeed, premium rates do not drop slowly when the cy-
cle heads downward. Rather, prices cascade downward, to
well below marginal cost. Industry Annual Statement operating

In truth, we lack information about expected profitability (and about expected duration
of profitability), and we lack good information about business strategies. Mathematical
expressions give an aura of empirical precision that is not warranted.
Perhaps one day we will have empirical data on the causes of underwriting cycles. We
do not have such data, and we do not pretend to have such data. This data provides an
intuitive understanding of underwriting cycles, based on types of market structures and
competitive strategies found in other industries.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [36] 09-12-02 2:43 pm

210 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

income was negative in 1975 and again in 1984–85. Moreover,

the reported operating ratios conceal the true severity of under-
writing cycles, for several reasons:
! First, accounting data does not include a “reasonable profit”
margin, although the economist’s marginal cost does. For in-
stance, a 2% accounting return on equity is a severe economic
loss.
! Second, most insurers desire steady earnings, particularly if
their financial statements are scrutinized by government regu-
lators or by stockholders. Insurers tend to under-reserve during
poor years, thereby increasing net income. Conversely, when
profits improve, insurers strengthen reserves of prior years,
dampening their reported earnings.
It is difficult to quantify these effects, since the “reason-
able insurance profit margin” is much disputed and reserve
strengthening and weakening is difficult to quantify. Never-
theless, rates were surely below marginal cost during 1974
and 1983 (in addition to 1975 and 1984–85).
! Third, the severity of the cycle differs by line. General Liabil-
ity rates, for example, were below marginal cost in 1982 and
perhaps in 1981 also. In other words, an accurate analysis of
income adjusted for reserve changes by line of business with
a reasonable profit provision shows severe price inadequacies
for several years in a row.
To recapitulate: during profitable years, there are incentives
for small firms to aggressively seek market share and for new
firms to enter the insurance industry. The lack of product dif-
ferentiation, the positive cash flow from insurance operations,
and the ease of entry would normally reduce or eliminate profits
from the industry.
Yet total consumer demand for insurance is inelastic with re-
spect to price. The difficulty of price comparisons and consumer
loyalty to insurers provide a large potential profit margin.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [37] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 211

The deciding factor is business strategy. If firms aggressively

seek market share by cutting rates, profits decline for all insurers.
Formal agreements to maintain high prices are not sustainable in
an industry as competitive as insurance. Rather, small firms and
new entrants may be dissuaded from pursuing overly aggressive
strategies by the competitive reactions of incumbent insurers.
Thus, the downward rate spiral is not a reflection of simple
competitive pricing. Rather, it is a competitive response to ag-
gressive strategies. By temporarily cutting rates below marginal
cost, incumbent insurers hope to persuade more aggressive but
short-sighted firms to modify their objectives from market share
to profitability.

Changing Strategies
Indeed, as operating profitability decreases, overly aggressive
insurers begin to rethink their strategy. First, low prices no longer
attract additional consumers, since even the major firms have cut
rates. Second, if profits remain negative, all firms suffer.
The changes in insurer strategies are revealed in the insurance
trade press and trade conferences. As the cycle deepens, laments
on the evils of price cutting become frequent, and exhortations to
refrain from the unprofitable pursuit of premium abound. These
public proclamations are disavowals of aggressive intentions. In-
surers say: “We renounce the use of rate reductions to gain mar-
ket share, for we see the folly of our ways.”
We can model the change in strategy as follows. As the trough
of the underwriting cycle continues, more firms renounce market
share gains and seek profitable business. The larger the expected
losses, and the longer the duration of the expected losses, the
more the firms emphasize increased profitability.

Industry Discipline
When the cycle turns up, insurers who previously engaged in
competitive “warfare” seem to raise rates in unison. Politicians,
 job no. 1987 casualty actuarial society CAS journal 1987D03 [38] 09-12-02 2:43 pm

212 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

consumer activists, and the legal community suspect antitrust vi-

olations. But there is no collusion, no intercompany agreements,
and only a general knowledge of competitors’ intentions.44
Rather, the change in behavior reflects the change in strategy.
The public exhortations during the trough of the cycle are not ac-
companied by rate increases. Each insurer knows that if it raises
prices unilaterally, it will lose business, not return to profitabil-
ity. In fact, most insurers always knew that severe rate cutting is
destructive to the industry. The public statements are intended to
persuade other firms to cease overly aggressive behavior. They
are not explanations of any firm’s current actions.45
Consider again the formal model. If economic profits are suf-
ficiently negative long enough, most firms will have shifted their
emphasis from market share growth to maintaining profitable
rates. Yet a high price maintenance strategy is profitable only if
all or most firms in the industry follow this path. Indeed, after
two or three years of pricing below marginal cost, most firms
are committed to writing profitable business. But how does one
move from a low price situation to a high price situation?

Market Leaders
In a highly competitive and fragmented industry like insur-
ance, firms cannot easily monitor the actions, much less the strat-
egies, of their rivals. They need a barometer of industry feelings.

44 See, for instance, the class action antitrust complaint in Van de Kamp [108] and an
industry response by the Insurance Information Institute [56].
45 Compare Porter [93, pg. 81]: “It is not uncommon for competitors to comment on
industry conditions : : : . Such commentary is laden with signals : : : . As such, this discus-
sion can be a conscious or unconscious attempt to get other firms to operate under the
same assumptions and thereby minimize the chance of mistaken motives and warfare.
Such commentary can also contain implicit pleas for price discipline: ‘Price competi-
tion is still very harsh. The industry is doing a lousy job of passing along increased
costs to the consumer.’ ‘The problem in this industry is that some firms do not recog-
nize that these current prices will be detrimental to our ability to grow and produce a
quality product in the long run.’ Or discussions of the industry may contain : : : implicit
promises to cooperate if others act ‘properly.’ ” [The quotations are from the president
of the Sherwin-Williams Coating Group and from an executive of a leading commodities
producer.]
 job no. 1987 casualty actuarial society CAS journal 1987D03 [39] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 213

Rate filings make dull newsprint. “The XYZ Insurance Com-

pany has requested a 5.1% rate increase in Arizona for Bodily
Injury coverage, 4.3% for Property Damage, : : : .” Who would
ever read such details?
The National Underwriter periodically records State Farm’s
rate filings (often only State Farm’s filings) in various jurisdic-
tions. State Farm is the market leader and low cost carrier in
Personal Lines coverages. It serves as the barometer of industry
movement through the underwriting cycle.46 By examining and
following State Farm’s actions, other firms maintain a close grasp
on industry price movements, even if they lack the resources to
monitor competitive rates on their own.
When other carriers see State Farm raising rates, they know
that firm strategies have shifted sufficiently to allow maintenance
of high prices. Insurers follow (or sometimes even anticipate) the
market leader in the various jurisdictions, leading to the good
years of the cycle.
In the Commercial Lines, there is no clear market leader. The
major Commercial Lines insurers, such as Travelers, Hartford,
CNA, AIG, and Liberty Mutual, have relatively small country-
wide market shares. Other carriers do not follow AIG’s Gen-
eral Liability rates the way they examine State Farm’s Personal
Auto rates. Consequently, the industry trade press rarely men-
tions Commercial Lines rate actions.47

46 Moreover, State Farm has a sophisticated monitoring system to analyze the rate actions
of its peer companies. Not only do State Farm’s rates affect a large percentage of the
insured population, but they also reflect of the strategies of other carriers.
47 Personal Lines risks are manually rated, so State Farm’s rate manual is an accurate
reflection of marketplace prices. Large Commercial Lines risks may be loss rated, com-
posite rated, schedule rated, or retrospectively rated. The rate manual is but a crude guide
to actual prices. In fact, many General Liability classifications are “A-rated,” so there are
no manual rates to examine.
In the Personal Lines, price changes are effected by rate filings. In the Commercial
Lines, prices may also be changed by varying schedule rating credits and debits, by
modifying the premium payment pattern, by changing policyholder dividend plans, and
by similar “non-manual” methods. Thus, rate comparisons are more difficult in the Com-
mercial Lines of business.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [40] 09-12-02 2:43 pm

214 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

Rating Bureaus
Rather, Commercial Lines insurance strategies are revealed
by deviations from bureau rates or bureau loss costs. The Na-
tional Council on Compensation Insurance (NCCI), and state
bureaus in certain jurisdictions (e.g., California, Massachusetts,
Minnesota, New York, Pennsylvania), provide loss costs for all
Workers Compensation classifications. Similarly, the Insurance
Services Office (ISO) provides loss cost data for the other Com-
mercial Lines. Most insurers use NCCI or ISO rates as a bench-
mark, and file rate deviations or independent rates with state
insurance departments.
After several years of unprofitable operations, insurers know
that the industry is ready to increase rates. ISO (or another rat-
ing bureau) provides the catalyst. When private insurers follow
ISO loss costs, without seeking major deviations, firms know
that the industry is committed to profitable rates. The individual
carriers may then curtail schedule rating credits and other price
modifications, confident that their rivals are doing the same.
Profits encourage aggressive rate cutting. After one or two
good years, insurer strategies begin emphasizing market share
growth, and new firms are attracted to the industry. The cycle
begins anew, in perpetual disequilibrium.

6. PUBLIC POLICY

As each cycle rolls through the industry, insurers ponder:

“What determines the severity and frequency of underwriting
cycles? What lines of business are most subject to them? When
will the cycle turn? How do state regulation and statutes influ-
ence cycles?” It is time to answer these questions.

Policyholder Loyalty and Price Elasticity

The beckoning of profits leads the cycle. Why drive rates
down if you cannot recoup the losses later? Firms would prefer
 job no. 1987 casualty actuarial society CAS journal 1987D03 [41] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 215

to price at marginal cost rather than lose money over the long
term.
Periods of high prices are sustainable only if consumers do
not reduce their purchases of the good and do not switch to rival
suppliers. In other words, the price elasticity of demand must be
low enough that consumer demand will not drop substantially
when suppliers raise prices.
Removing statutory requirements for Personal Automobile
and Workers Compensation insurance, and curtailing judicial
awards in commercial liability cases, might increase the price
elasticity of demand for insurance. But the statutory insurance
requirements help the victims of motor vehicle and workplace
accidents. The benefits they provide outweigh the disadvantages
of premium rate fluctuations.
The unpredictability of jury awards in commercial liability
cases provides little social benefit, and the harm to society ex-
tends beyond insurance availability and rate fluctuation concerns.
Unfortunately, the limited success of tort reform efforts in the
1980s and early 1990s highlights the intractability of this prob-
lem. To restate this: the trial bar is a powerful interest group
that opposes tort reform. The results of the pervasive attorney
involvement in insurance claims are bloated insurance costs and
the redistribution of wealth from citizens to a particular pro-
fession (AIRAC [2; 3]). More volatile underwriting cycles are
simply an additional side-effect.
Policyholder loyalty results from the difficulty of price com-
parisons. Personal Lines policyholders may be unaware of price
slashing by competing insurers, since they rarely price shop at
renewal. An insurer can maintain high prices for a short period
without a major loss of market share when its competitors begin
cutting rates.
Price increases, however, encourage insureds to seek better
rates elsewhere. Unilateral price increases cause a loss of market
share, as consumers switch to rival carriers. Industry-wide price
 job no. 1987 casualty actuarial society CAS journal 1987D03 [42] 09-12-02 2:43 pm

216 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

increases are easier to sustain, since consumers cannot do better

elsewhere in the marketplace. Thus, the descent to the trough of
the cycle is precipitated by a small group of insurers, but the
return to profitability is a uniform movement.
Greater consumer price information would reduce loyalty to
the current insurer and mitigate the severity of underwriting cy-
cles.48 Firms would not be able to sustain high prices in the
face of competitive price cutting without rapidly losing market
share. Prices closer to cost would prevail over the duration of the
underwriting cycle.
Life insurance regulation demonstrates the difficulty of pro-
viding price comparisons. The NAIC Life Insurance Solicitation
Model Regulation requires that insurers illustrate surrender cost
and net payment cost indices for 10 and 20 year durations, but
few consumers examine these numbers (Black and Skipper [15]).
Such comparisons are difficult, and few individuals expend the
effort to understand them.
The same is true for Property/Casualty insurance. Consumers
do not forgo price comparisons because the information is not
available. Rather, the information is not available because the
price comparisons are so distasteful.

Underwriting Cycles by Line

The history of underwriting cycles in America illustrates these
relationships (see Figures 2 and 3). During the 1960s and 1970s,
underwriting cycles were most pronounced for Personal Auto-
mobile and Workers Compensation insurance.49 In the 1980s,

48 Numerous studies have recommended that states make insurance price information
accessible to consumers; see Virginia Bureau of Insurance [114], recommendation #5,
or NAIC [81, pp. 440–441].
49 See Stewart [101, Exhibits 5-3, 5-4, and 5-8 on pp. 290, 291, and 295]. Note how
the cycles in automobile insurance mirrored those for the industry as a whole, whereas
General Liability showed no clear pattern until the late 1970s. Similarly, Best’s [12, pg.
33] notes that “while the majority of insolvencies during the 1970s occurred in personal
lines companies, commercial lines companies accounted for the majority in the 1980s.”
job no. 1987 casualty actuarial society CAS journal 1987D03 [43] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 217

 job no. 1987 casualty actuarial society CAS journal 1987D03 [44] 09-12-02 2:43 pm

218 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

 job no. 1987 casualty actuarial society CAS journal 1987D03 [45] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 219

General Liability and other Commercial Liability lines showed

the greatest fluctuations in profitability: negative in 1981–1984
and highly positive in 1986–1988.
This difference is influenced by demand elasticities and bar-
riers to entry. Personal Automobile and Workers Compensation
insurance are statutorily mandated by Financial Responsibility or
compulsory insurance laws. Price elasticity of demand is low.50
The opposite was true for General Liability until the 1970s.
Believing that they had little exposure to liability hazards, many
small businesses declined to purchase the coverage. Large cor-
porations often used other risk management techniques, such as
self-funding and captives.
In the 1950s and 1960s, many Personal Lines insurers used
bureau rates, either as actual rates or as a baseline for pricing.
By the 1980s, the low cost direct writers, such as State Farm
and Allstate, had garnered most of the Personal Lines market.
The efficient distribution systems of these insurers formed strong
barriers to entry or expansion by other firms.
The opposite course has characterized the Commercial Liabil-
ity lines of business. The major direct writers do not dominate
these markets. Moreover, the lengthening tails in these lines and
the rising interest rates in the 1970s increased the disparity be-
tween bureau rates and marginal cost.

50 On the low price elasticity of demand, see Sherdan [99, pg. 58]; Bloom [16]; and
Strain [103, pg. 448]. Strain summarizes the influences on elasticity as “The greater
the tendency for the public to buy an insurance coverage without the need for sales
stimulation (as to comply with financial responsibility laws, or workmen’s compensation
acts, or mortgage protection requirements), the more inelastic the demand for insurance.”
Financial Responsibility laws require a driver involved in a motor vehicle accident either
to show evidence of insurance or to post a court bond (Morill [77]; Mehr and Cammack
[72, pp. 308–329]; Bickelhaupt [13, pp. 646–678]). Employers must provide Workers
Compensation insurance, with minor exceptions that are relating to farm employment,
household work, or businesses with few workers. Employers that are financially strong
enough to self-insure may provide the statutory benefits on their own. For history and
detail, see Myers [79, pp. 884–900], Kulp and Hall [64, pp. 191–250], and Chamber of
Commerce [25]. Many states allow group self-insurance (NAIC [82]). This increases the
price elasticity of demand, since consumers have another risk management technique.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [46] 09-12-02 2:43 pm

220 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

The high costs of Workers Compensation insurance in the

late 1980s, exacerbated by large residual market loads in many
jurisdictions, led many employers to alternative risk manage-
ment techniques, such as group self-insurance and large dollar
deductible policies. Price elasticity of demand increased, and a
uniform increase in price would drive the better risks from the
insurance market. So Workers Compensation remained unprof-
itable through the 1980s, until the state legislative reforms and
the managed care revolution of the 1990s lowered loss costs
without necessitating large rate increases.

Regulation and Social Developments

Changes in state regulation may influence underwriting cy-
cles. During the 1960s and early 1970s, many states moved from
prior approval regulation to open competition laws.51 Competi-
tive rating laws allow more freedom for private insurers to vary
premium rates in attempts to gain market share or increase prof-
its.
The 1980s and 1990s show ambiguous trends. California
adopted prior approval regulation in November 1988, with the
passage of Proposition 103, and consumer groups in other states
are pushing similar legislation. Meanwhile, the low cost direct
writers are driving agency companies out of the Personal Lines
market. Tighter governmental regulation and increasing market
concentration may dampen the severity of Personal Automobile
underwriting cycles.52
Social developments in the 1980s and 1990s have had the
opposite effect on the Commercial Liability lines. The expansion
of tort law doctrines, and the increasing unpredictability of jury
awards, have made coverage essential even for small firms. State

51 See NAIC [83, pg. 310]: “It is the sense of the Subcommittee : : : that : :: reliance be
placed upon fair and open competition to produce and maintain reasonable and compet-
itive prices for insurance coverages : : : .” See also DOJ [41]).
52 Compare Eley [43, pg. 187]: “If the likelihood of extraordinary profits during hard mar-
kets is removed, the willingness of insurers to give away insurance during soft markets
will evaporate.”
 job no. 1987 casualty actuarial society CAS journal 1987D03 [47] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 221

regulation is less restrictive, since commercial insureds can fend

for themselves and do not need the governmental protection that
ordinary citizens require. The major rating bureaus, such as ISO
and NCCI, have changed from advisory rates to loss costs in
most jurisdictions, and may soon be further transformed into
quasi-consulting organizations. Commercial lines rate and form
deregulation is possible in the early years of the 21st century.
Consequently, General Liability promises potential profits for
the discerning insurer.53 In the late 1970s, insurers complained
vociferously about rising and unjustified liability awards. The
criticism was correct: the American legal system encourages law-
suits and the redistribution of wealth from the public to the trial
bar. But a secondary effect of these complaints was to impress
upon businesses the need for liability coverage.
Numerous suppliers—major carriers, small firms, and new
entrants—joined the fray, and insurers began positioning them-
selves (that is, cutting prices to build market share) for the antici-
pated profits. The aggressive competition threatened to eliminate
the foreseen returns.
So General Liability entered the trough of a severe under-
writing cycle, with firms slashing rates well below cost. The
consequences were striking: when rates rose in 1985, there was
an almost complete absence of aggressive price cutting.54

53 This promise may prove illusory. Insurers who provided CGL coverage in the 1960s
and 1970s are now facing enormous asbestos, pollution, and products liability litigation
(Hamilton and Routman [52]; Manta and Welge [69]). Nevertheless, the potential is
alluring.
54 The power of underwriting cycles is often misunderstood. Much of the American legal
community and the business public concluded that the dramatic and uniform rise in
Commercial Liability insurance rates must be the result of collusion. Yet no evidence of
such behavior could be found. In fact, collusion is nearly impossible in the fragmented
insurance market.
Even the Attorneys General’s antitrust complaint was confined to allegations of boy-
cott in policy form development, statistical support, and coverage exclusions. Pricing in
concert is never mentioned (Van de Kamp [108]). The California Attorney General’s
office explains that pricing in concert is protected by the McCarran-Ferguson Act and so
was not contested. An alternative explanation is that the Commercial Liability insurance
rate increases were characterized not by pricing in concert but by the competition driving
the underwriting cycle.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [48] 09-12-02 2:43 pm

222 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

And the cycle continues. The aggressive competition that pre-

cipitated the rise in rates in the mid-1980s led to price cutting a
few years later. The waning influence of rating bureaus and ad-
ministered pricing systems in the fragmented insurance market
will lead to even more severe swings in premiums.

7. CONCLUSION
Underwriting cycles are a means of maintaining long-term
profits, not a random occurrence that removes them. Insurance
underwriting cycles are the display of competitive pricing in a
free marketplace. To optimize the results of their companies,
pricing actuaries must learn to adapt their rate setting techniques
to the phases of the underwriting cycle.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [49] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 223

REFERENCES AND BIBLIOGRAPHY

[1] AICPA, Insurance Companies Committee, Auditing Stan-

dards Division, “Discounting Claims of Insurance Enter-
prises,” Draft Issues Paper, March 1983.
[2] All-Industry Research Advisory Council, Attorney Involve-
ment in Auto Injury Claims, Oak Brook, IL: AIRAC, 1988.
[3] All-Industry Research Advisory Council, Compensation
for Automobile Injuries in the United States, Oak Brook,
IL: AIRAC, March 1989.
[4] Almagro, Manuel, and Thomas L. Ghezzi, “Federal In-
come Taxes—Provisions Affecting Property/Casualty In-
surers,” PCAS LXXV, 1988, pp. 95–161.
[5] Anderson, Dan R., and Roger A. Formisano, “Causal Fac-
tors in P-L Insolvency,” Journal of Insurance Regulation,
Volume 6, No. 4 (June 1988), pp. 449–461.
[6] Appel, David, and James Gerofsky, “Regulating Compe-
tition: The Case of Workers Compensation Insurance,”
Journal of Insurance Regulation, Volume 3, No. 4 (June
1985), pp. 408–425.
[7] Appel, David, and James Gerofsky, “Regulating Compe-
tition: The Case of Workers Compensation Insurance,”
NCCI Digest, Volume 2, Issue 3 (September 1987), pp.
1–18.
[8] Balcarek, Rafal J., “The Capital Investment Market and
the Insurance Industry,” PCAS LV, 1968, pp. 186–195.
[9] Banfield, Carole, “P/C Insurer Profitability: ISO Paid
Claims Data,” Journal of Insurance Regulation, Volume 5,
No. 2 (December 1986), pp. 268–276.
[10] Bailey, Robert A., “A Review of the Little Report on Rates
of Return in the Property and Liability Insurance Indus-
try,” PCAS LVI, 1969, pp. 133–162.
[11] Berger, Lawrence, “A Model of the Underwriting Cycle
in the Property and Casualty Insurance Industry,” Journal
of Risk and Insurance, Volume 55 (1988).
 job no. 1987 casualty actuarial society CAS journal 1987D03 [50] 09-12-02 2:43 pm

224 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

[12] Best’s Insolvency Study: Property/Casualty Insurers 1969–

1990, Oldwick, NJ: A. M. Best Company, June 1991.
[13] Bickelhaupt, David L., General Insurance, Tenth Edition,
Homewood, IL: Richard D. Irwin, Inc., 1979.
[14] Bierwag, G. O., George G. Kaufman, and Alden Toevs,
“Duration: Its Development and Use in Bond Portfolio
Management,” Financial Analysts Journal, July–August
1983.
[15] Black, Kenneth, Jr., and Harold D. Skipper, Jr., Life In-
surance, Twelfth Edition, Englewood Cliffs, NJ: Prentice-
Hall, 1994.
[16] Bloom, Thomas S., “Cycles and Solutions,” Best’s Re-
view: Property/Casualty Edition, Volume 86, No. 6 (Oc-
tober 1987), pp. 20 ff.
[17] Boor, Joseph, “A Macroeconomic View of the Insurance
Marketplace,” CAS Exam 5 Study Note, 1999A.
[18] Boor, Joseph, “The Impact of the Insurance Economic
Cycle on Insurance Pricing,” CAS Exam 5 Study Note,
1999B.
[19] Braithwaite, Paul, “P/C Profitability: The Long-Term
View,” Journal of Insurance Regulation, Volume 6, No.
1 (September 1987), pp. 96–104.
[20] Brock, Gerald W., “The Computer Industry,” in Walter
Adams (ed.), The Structure of American Industry, New
York: MacMillan Publishing Company, 1986, pp. 239–
260.
[21] Butsic, Robert P., “Determining the Proper Interest Rate
for Loss Reserve Discounting: An Economic Approach,”
Evaluating Insurance Company Liabilities, Casualty Actu-
arial Society Discussion Paper Program, 1988, pp. 147–
188.
[22] Butsic, Robert P., “The Effect of Inflation on Losses and
Premiums for Property-Liability Insurers,” Inflation Impli-
cations for Property-Casualty Insurance, Casualty Actuar-
ial Society Discussion Paper Program, 1981, pp. 51–102.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [51] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 225

[23] Butsic, Robert P., “The Underwriting Cycle: A Necessary

Evil?” The Actuarial Digest, Volume 8, No. 2 (April/May
1989).
[24] Carlson, Thomas O., “Rate Regulation and the Casualty
Actuary,” PCAS XXXVIII, 1951, pp. 9–72; discussions
by H. J. Ginsburgh, pp. 222–223; Allie V. Resony, pp.
223–225; Dudley M. Pruitt, pp. 225–230; R. W. Griffith,
pp. 230–236; author’s review of discussions, pp. 236–239.
[25] Chamber of Commerce of the United States, Analysis
of Workers’ Compensation Laws, Washington, D.C.: U.S.
Chamber of Commerce, 1990.
[26] Cholnoky, Thomas V., and Jeffrey Cohen, Property/Cas-
ualty 10K Loss Reserve Disclosures, Goldman Sachs, May
31, 1989.
[27] Conning & Co., New Business versus Renewals: The Cost
of New Business in a Soft Market, Hartford, CT: Conning
& Co., June 1988.
[28] Conning and Company, The Real Property & Casualty Cy-
cle: Behind the Reported Results, Conning & Co., Decem-
ber 1987.
[29] Countryman, Gary L., “Open Competition Rating for
Workers’ Compensation,” Journal of Insurance Regulation,
Volume 1, No. 1 (September 1982), pp. 9–22.
[30] Cummins, J. David, and Scott E. Harrington, “Property-
Liability Insurance Rate Regulation: Estimation of Under-
writing Betas Using Quarterly Profit Data,” The Journal
of Risk and Insurance, Volume 52, No. 1 (March 1985),
pp. 16–43.
[31] Cummins, J. David, Scott E. Harrington, and Robert
W. Klein, “Cycles and Crises,” Best’s Review: Prop-
erty/Casualty Edition, Volume 92, No. 9 (January 1992),
pp. 15 ff.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [52] 09-12-02 2:43 pm

226 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

[32] Cummins, J. David, Scott E. Harrington, and Robert W.

Klein, “Cycles and Crises in Property/Casualty Insurance:
Causes and Implications for Public Policy,” Journal of In-
surance Regulation, Volume 10, No. 1 (Fall 1991), pp.
50–93.
[33] Cummins, J. David, Scott E. Harrington, and Robert W.
Klein (eds.), Cycles and Crises in Property/Casualty Insur-
ance: Causes and Implications for Public Policy, National
Association of Insurance Commissioners, 1991.
[34] Cummins, J. David, and J. François Outreville, “An In-
ternational Analysis of Underwriting Cycles in Property-
Liability Insurance,” Journal of Risk and Insurance, Vol-
ume 54, Number 2 (June 1987), pp. 246–252.
[35] Cummins, J. David, and David J. Nye, “Inflation and
Property-Liability Insurance,” in John D. Long and Ev-
erett D. Randall (eds.) Issues in Insurance, Volume I, Third
Edition, Malvern, PA: American Institute for Property and
Liability Underwriters, 1984.
[36] Danzon, Patricia Munch, “Rating Bureaus in U.S. Property
Liability Insurance Markets: Anti or Pro-competitive?”
The Geneva Papers on Risk and Insurance, Vol 8, No. 29
(October 1983), pp. 371–402.
[37] D’Arcy, Stephen P., “Application of Economic Theories of
Regulation to the Property-Liability Insurance Industry,”
Journal of Risk and Insurance, Volume 7, No. 1 (September
1988), pp. 19–51.
[38] D’Arcy, Stephen P., and Neil A. Doherty, The Financial
Theory of Pricing Property-Liability Insurance Contracts,
Homewood, IL: Richard D. Irwin, Inc., 1988.
[39] D’Arcy, Stephen P., and James R. Garven, “Property-
Liability Insurance Pricing Models: An Empirical Eval-
uation,” Journal of Risk and Insurance, Volume 57, No. 3
(September 1990), pp. 391–419.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [53] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 227

[40] Daykin, Chris D., Teivo Pentikäinen, and M. Pesonen,

Practical Risk Theory for Actuaries, First Edition, London:
Chapman and Hall, 1994.
[41] Department of Justice, “Report to the Task Force on An-
titrust Immunities,” The Pricing and Marketing of Insur-
ance, January 1977.
[42] Douglas, Evan J., Intermediate Microeconomic Analysis:
Theory and Applications, Englewood Cliffs, NJ: Prentice-
Hall, Inc., 1982.
[43] Eley, David F., “Investment Income in Ratemaking: An
Alternative to the Cycle,” Journal of Insurance Regulation,
Volume 5, No. 2 (December 1986), pp. 180–188.
[44] Ezekiel, Mordechai, “The Cobweb Theorem,” Quarterly
Journal of Economics (February 1938), pp. 255–280;
reprinted in Haberler, Gottfried, (ed.), Readings in Busi-
ness Cycle Theory, Philadelphia: The Blakiston Company,
1944, pp. 422–442.
[45] Fairley, William, “Investment Income and Profit Margins
in Property-Liability Insurance: Theory and Empirical Re-
sults,” The Bell Journal of Economics 10 (Spring 1979),
pp. 192–210; reprinted in J. David Cummins and Scott E.
Harrington (eds.), Fair Rate of Return in Property-Liability
Insurance, Boston: Kluwer-Nijhoff Publishing, 1987, pp.
1–26.
[46] Feldblum, Sholom, “Asset Liability Matching for Prop-
erty/Casualty Insurers,” Valuation Issues, Casualty Actu-
arial Society Discussion Paper Program, 1989, pp. 117–
154.
[47] Fox, Eugene H., “Automobile Insurance Analysis: The
Twenty-Five and Under Age Driver vs. the Twenty-Six
and Over Age Driver,” CPCU Journal, Volume 33, No. 1
(March 1980), pp. 22–25.
[48] Friedman, James, Oligopoly Theory, Cambridge: Cam-
bridge University Press, 1983.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [54] 09-12-02 2:43 pm

228 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

[49] Fudenberg, Drew, and Jean Tirole, “Noncooperative Game

Theory for Industrial Organization: An Introduction and
Overview,” in Richard Schmalensee and Robert D. Willig
(eds.), Handbook of Industrial Organization, North-Hol-
land: Elsevier Science Publishing Co., 1989, Volume I,
pp. 159–328.
[50] Gleeson, Owen M., and Gerald I. Lenrow, “An Analy-
sis of the Impact of the Tax Reform Act on the Prop-
erty/Casualty Industry,” in Financial Analysis of Insurance
Companies, Casualty Actuarial Society Discussion Paper
Program, 1987, pp. 119–190.
[51] Greene, Winfield W., and William F. Roeber, “The ‘Per-
manent’ Rate Making Method Adopted by the National
Council on Compensation Insurance,” PCAS XII, 1925–
26, pp. 253–267.
[52] Hamilton, Thomas M., and Eric L. Routman, “Cleaning
Up America: Superfund and Its Impact on the Insurance
Industry,” CPCU Journal, Volume 41, No. 3 (September
1988), pp. 172–184.
[53] Harrington, Scott, “The Impact of Insurance Regulation on
Prices and Underwriting Results in the Property-Liability
Insurance Industry: A Survey,” Journal of Risk and Insur-
ance, Volume 51, No. 4 (December 1984), pp. 577–623.
[54] Harrington, Scott E., and Patricia M. Danzon, “Price
Cutting and Premium Volatility in Liability Insurance
Markets,” in J. David Cummins, Scott E. Harrington,
and Robert W. Klein (eds.), Cycles and Crises in Prop-
erty/Casualty Insurance: Causes and Implications for Pub-
lic Policy, National Association of Insurance Commission-
ers, 1991.
[55] Hill, Raymond, “Profit Regulation in Property-Liability
Insurance,” The Bell Journal of Economics, Vol. 10, No. 1
(Spring 1979), pp. 172–191.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [55] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 229

[56] Insurance Information Institute, Insurance Antitrust Liti-

gation: The Conspiracy Theory and Related Policy Issues,
April 1988.
[57] Insurance Services Office, Changes in the Economic Envi-
ronment: Insurance Implications, October 1991.
[58] Joskow, Paul L., “Cartels, Competition, and Regulation in
the Property-Liability Insurance Industry,” The Bell Jour-
nal of Economics and Management Science, Volume 4, No.
2 (Autumn 1973), pp. 375–427.
[59] Joskow, Paul, “Competition and Regulation in the Prop-
erty/Casualty Insurance Industry,” Report of the Economic
Advisory Panel to the National Commission for the Review
of Antitrust Laws and Procedures, pp. 207–227.
[60] Kahane, Yehuda, “Generation of Investable Funds and the
Portfolio Behavior of the Non-Life Insurers,” Journal of
Risk and Insurance, Volume 45, No. 1 (March 1978), pp.
65–77.
[61] Klein, Robert, “Competition in Private Passenger Auto-
mobile Insurance,” A Report to the NAIC Personal Lines
(C) Committee, September 1989.
[62] Klein, Robert W., “Competitive Rating in Workers’ Com-
pensation Insurance: Michigan’s Experience,” Journal of
Insurance Regulation, Volume 5, No. 1 (September 1986),
pp. 79–108.
[63] Kozik, Thomas J., “Underwriting Betas—The Shadows of
Ghosts,” PCAS LXXXI, 1994, pp. 303–329.
[64] Kulp, C. A., and John W. Hall, Casualty Insurance, Fourth
Edition, New York: John Wiley & Sons, 1968.
[65] Longley-Cook, Laurence H., “Early Actuarial Studies in
the Field of Property and Liability Insurance,” PCAS LI,
1964, pp. 140–146.
[66] Longley-Cook, Laurence H., “Underwriting Profit in Fire
Bureau Rates,” PCAS LII, 1966, pp. 305–311.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [56] 09-12-02 2:43 pm

230 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

[67] Lowe, Stephen P., “A New Performance Measure for P/C

Insurers,” Emphasis, Summer 1988, pp. 8–11.
[68] Mahler, Howard C., “An Introduction to Underwriting
Profit Models,” PCAS LXXI, 1987, pp. 239–277.
[69] Manta, Joseph G., and Mark A. Welge, Toxic Tort—
Environmental Hazards Litigation Workshop, Philadelphia,
1990.
[70] Matison, Raymond A., “Capital Cornucopia,” Best’s Re-
view: Property/Casualty Edition, Volume 88, No. 7 (Nov-
ember 1987), pp. 26 ff.
[71] McGee, Robert T., “The Cycle in Property/Casualty Insur-
ance,” Federal Reserve Bank of New York Quarterly Review,
Autumn 1986, pp. 22–30.
[72] Mehr, Robert I., and Emerson Cammack, Principles of In-
surance, Seventh Edition, Homewood, IL: Richard D. Ir-
win, Inc., 1980.
[73] Meidan, Arthur, “Marketing Strategies, Organization and
Performance Control in Insurance,” Journal of Risk and
Insurance, Volume 49, No. 3 (September 1982), pp. 388–
404.
[74] Meyerson, Allen L., “Ensuring the Solvency of Property
and Liability Insurance Companies,” in Spencer L. Kim-
ball and Herbert S. Denenberg (eds.), Insurance, Govern-
ment, and Social Policy, Homewood, IL: Richard D. Irwin,
Inc., 1969, pp. 146–190.
[75] Miccolis, Robert S., “On the Theory of Increased Limits
and Excess of Loss Pricing,” PCAS LXIV, 1977, pg. 27.
[76] Milgrom, Paul, and John Roberts, “Limit Pricing and En-
try Under Incomplete Information: An Equilibrium Anal-
ysis,” Econometrica, Volume 50 (1982), pp. 443–460.
[77] Morrill, Thomas C., “Motorists’ Financial Responsibility
Laws,” in John D. Long and Davis W. Gregg (eds.), Prop-
erty and Liability Insurance Handbook, Homewood, IL:
Richard D. Irwin, 1965, pp. 576–599.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [57] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 231

[78] Mowbray, Albert H., and S. B. Black, “On the Relation

of Accident Frequency to Business Activity,” PCAS II,
1915–1916, pp. 418–426.
[79] Myers, Robert J., Social Security, Third Edition, Home-
wood, IL: Richard D. Irwin, Inc., 1985.
[80] Myers, Stewart C., and Richard A. Cohn, “A Discounted
Cash Flow Approach to Property-Liability Insurance Rate
Regulation,” in Cummins and Harrington (eds.), Fair Rate
of Return in Property-Liability Insurance, Boston: Kluwer-
Nijhoff Publishing, 1987, pp. 55–78.
[81] National Association of Insurance Commissioners, “Re-
port of the Advisory Committee on Competitive Rating to
the National Association of Insurance Commissioners,”
Proceedings of the National Association of Insurance Com-
missioners, 1980, Volume II, pp. 440–441.
[82] National Association of Insurance Commissioners Study
Committee Report, “Regulating Workers’ Compensation
Groups,” in John D. Long and Everett D. Randall (eds.),
Issues in Insurance, Volume II, Third Edition, Malvern,
PA: The American Institute for Property and Liability In-
surance, 1984, pp. 1–110.
[83] National Association of Insurance Commissioners, Pro-
ceedings, 1969, pg. 310.
[84] Nash, John F., Jr., “Noncooperative Games,” Annals of
Mathematics, Volume 45 (1951), pp. 286–295.
[85] Nelson, Robert E., “Property-Liability Company Exits,”
Journal of Risk and Insurance, Volume 38, No. 3 (Septem-
ber 1971), pp. 357–366.
[86] Outreville, J. François, “Underwriting Cycles and Rate
Regulation in Automobile Insurance Markets,” Journal of
Insurance Regulation, Volume 9, No. 2 (December 1990),
pp. 274–286.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [58] 09-12-02 2:43 pm

232 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

[87] Panning, William H., “Asset-Liability Matching for a Go-

ing Concern,” in Edward I. Altman and Irwin T. Van-
derhoof (eds.), The Financial Dynamics of the Insurance
Industry, New York: Irwin, 1995, pp. 257–292.
[88] Pentikäinen, Bonsdorff, Pesonen, Rantala, and Ruoho-
nen, Insurance Solvency and Financial Strength, Helsinki,
Finland: Finnish Insurance Training and Publishing Co.,
1989.
[89] Plotkin, Irving H., Prices and Profits in the Property and
Liability Insurance Industry, report to the American Insur-
ance Association, Cambridge, MA: Arthur D. Little, Inc.,
November 1967.
[90] Plotkin, Irving H., “Rates of Return in the Property and
Liability Insurance Industry: A Comparative Analysis,”
Journal of Risk and Insurance, Volume 36, No 3 (June
1969), pp. 173–200.
[91] Plotkin, Irving H., Studies on the Profitability, Industrial
Structure, Finance, and Solvency of the Property and Liabil-
ity Insurance Industry, Cambridge, MA: Arthur D. Little,
1970.
[92] Plotkin, Irving H., “Total Rate of Return and the Reg-
ulation of Insurance Profits,” Total Return Due a Prop-
erty/Casualty Insurance Company, Casualty Actuarial So-
ciety Discussion Paper Program, 1979, pp. 206–245.
[93] Porter, Michael, Competitive Strategy, New York: The Free
Press, 1980.
[94] Redington, F. M., “Review of the Principles of Life-Office
Valuations,” Journal of the Institute of Actuaries, Volume
18 (1952), pp. 286–340.
[95] Resony, Allie V., “Discussion of Thomas O. Carlson’s
‘Rate Regulation and the Casualty Actuary,’ ” PCAS
XXXVIII, 1951, pp. 223–225.
[96] Ryan, Kevin M., and Richard I. Fein, “A Forecast for
Workers Compensation,” NCCI Digest, Volume III, Issue
IV (December 1988).
 job no. 1987 casualty actuarial society CAS journal 1987D03 [59] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 233

[97] Scherer, Frederick M., Industrial Market Structure and Eco-

nomic Performance, Second Edition, Boston: Houghton
Mifflin Company, 1980.
[98] Shapiro, Carl, “Theories of Oligopoly Behavior,” in
Richard Schmalensee and Robert D. Willig (eds.), Hand-
book of Industrial Organization, North-Holland: Elsevier
Science Publishing Co., 1989, Volume I, pp. 329–414.
[99] Sherdan, William A., “An Analysis of the Determinants of
Demand for Automobile Insurance,” The Journal of Risk
and Insurance, Volume 51, No. 1 (March 1984), pp. 49–
62.
[100] Stern, Lawrence J., “Corporate Changes 1989,” Best’s
Review: Property/Casualty Edition, Volume 90, No. 11
(March 1990), pp. 20–22.
[101] Stewart, Barbara D., “Profit Cycles in Property-Liability
Insurance,” in John D. Long and Everett D. Randall (eds.),
Issues in Insurance, Volume I, Third Edition, Malvern, PA:
American Institute for Property and Liability Underwrit-
ers, 1984, pp. 273–334.
[102] Stone, James M., “A Theory of Capacity and the Insurance
of Catastrophe Risks,” The Journal of Risk and Insurance,
Part I in Volume 40, Number 2 (June 1973), pp. 231–243,
and Part II in Volume 40, Number 3 (September 1973),
pp. 339–355.
[103] Strain, Robert W., “Insurance Pricing and Its Role in Re-
lation to Economic Theory and Marketing Management,”
Journal of Risk and Insurance, Volume 33, No. 3 (Septem-
ber 1966), pp. 447–457.
[104] Tarbell, Thomas F., “The Effect of Changes in Values on
Casualty Insurance,” PCAS XIX, Part 1, 1932–33, pp. 1–
12.
[105] Taylor, Gregory C., “Underwriting Cycles,” Coopers and
Lybrand, 1991.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [60] 09-12-02 2:43 pm

234 UNDERWRITING CYCLES AND BUSINESS STRATEGIES

[106] Tennyson, Sharon, “The Effect of Rate Regulation on Un-

derwriting Cycles,” CPCU Journal, Volume 44, No. 1
(March 1991), pp. 33–45.
[107] Tirole, Jean, The Theory of Industrial Organization, Cam-
bridge, MA: The MIT Press, 1988.
[108] Van de Kamp, John K., et al., The State of California ver-
sus Hartford Fire Insurance Company, et al., The United
States District Court for the Northern District of Califor-
nia, March 1988.
[109] Varian, Hal R., Intermediate Microeconomics: A Modern
Approach, New York: W. W. Norton and Company, 1987.
[110] Varian, Hal R., Microeconomic Analysis, Second Edition,
New York: W. W. Norton and Company, 1984.
[111] Venezian, Emilio C., “Ratemaking Methods and Profit Cy-
cles in Property and Liability Insurance,” Journal of Risk
and Insurance, Volume 52, Number 3 (September 1985),
pp. 477–500.
[112] Victor, Richard A., “Major Challenges Facing Workers’
Compensation Systems in the 1990s,” in Richard A. Vic-
tor (ed.), Challenges for the 1990s, Cambridge, MA: Work-
ers Compensation Research Institute, July 1990, pp. 9–23.
[113] Victor, Richard B., and Charles A. Fleischman, How
Choice of Provider and Recessions Affect Medical Costs in
Workers’ Compensation, Cambridge, MA: Workers Com-
pensation Research Institute, 1990.
[114] Virginia Bureau of Insurance, State Corporation Commis-
sion, Competition in the Property and Casualty Insurance
Industry: An Evaluation of Alternative Methods of Rate Reg-
ulation, Richmond, Virginia: January 1978.
[115] Waterson, Michael, Economic Theory of the Industry, Cam-
bridge: Cambridge University Press, 1984.
[116] Woll, Richard G., “Insurance Profits: Keeping Score,” Fi-
nancial Analysis of Insurance Companies, Casualty Actuar-
ial Society Discussion Paper Program, 1987, pp. 446–533.
 job no. 1987 casualty actuarial society CAS journal 1987D03 [61] 09-12-02 2:43 pm

UNDERWRITING CYCLES AND BUSINESS STRATEGIES 235

[117] The Workers’ Compensation Rating and Inspection Bu-

reau of Massachusetts, 1977 Rate Filing, “Expense Cash
Flow” section.