initial scenario
FIGURE 5
FIG#
actual expected
cost per policy
$600
$800
$1,000
45
6
7
FIGURE 6
FIG#
5
6
7
FIGURE 1
FIG#
1
2
3
45
6
7
FIGURE 2
FIG#
laggard’s Book of Business
policy premium
(average cost +
10%)
$880
$880
$880
profit
$280
$80
–$120
$240
luminary’s Book of Business
policy premium
(accurate cost +
10%)
$660
$880
$1,100
actual expected
cost per policy
$600
$800
$1,000
total
1
2
3
45
6
7
FIGURE 3
FIG#
total
profit
$60
$80
$100
$240
FIGURE 4
after one year
FIGURE 4
FIG#
FIGURE 5
FIG#
FIGURE 6
FIG#
actual expected
cost per policy
$800
$1,000
$1,000
5
6
7
FIGURE 7
FIG#
laggard’s Book of Business
policy premium
(average cost +
10%)
$880
$880
$880
total
profit
$80
–$120
–$120
–$160
Because Laggard lost its lower-cost policy, it has lost the expected profit from that policy. Meanwhile, Luminary has gained
an additional $60 in profit from it. Luminary has lost the profit
it could have earned on the higher-cost policy, but this policy
compounds the loss that Laggard is now experiencing.
The cycle continues (see Figure 5), as Laggard must increase
its policy premiums to $1,027 to maintain its profit expectations. Now, Laggard’s $800 policy will be lost to the competitor,
since Luminary is able to offer a lower premium with its more
refined rate plan. After this policy is lost, Laggard’s book has
shrunk to only two high-cost policies, while Luminary now has
four policies—all of which are profitable.
The final iteration of the adverse-selection cycle occurs when
Laggard adjusts its policy premium one more time to reach its
profit expectations (see Figure 6). At this point, Laggard now
charges the same amount as Luminary would charge for a $1,000
cost policy. Laggard and Luminary compete for these higher-cost
policies, so each company will retain one of them.
In the real world, where there’s a continuum of higher- and
lower-cost policies, this cycle of adverse selection would continue. The lower-cost policies will be attracted to the company
with the more refined rate plan, while the unrefined rate plan
increasingly attracts only the highest-cost policies. But even
in this abbreviated example, the cycle of adverse selection has
luminary’s Book of Business
policy premium
(accurate cost +
10%)
$660
$660
$880
actual expected
cost per policy
$600
$600
$800
total
profit
$60
$60
$80
$200
left Laggard with a diminished book of business with much less
revenue and profit. Because Laggard doesn’t have the ability to
refine its loss cost estimates, it cannot compete for lower-cost
policies. In fact, Laggard may not even be able to cover its fixed
costs anymore and certainly will not be able to meet its shareholders’ growth expectations. Luminary has gained significant
advantage over Laggard because of its early adoption of the new
predictive information.
This simple example helps illustrate how the introduction of new predictive information can fuel competition in
the insurance marketplace. Those companies that refine
their rate plans gain a crucial competitive advantage, while
those that are slow to adopt are doomed to suffer the recurring onslaught of adverse selection. When new predictive
information enters the marketplace, there’s no such thing as
maintaining the status quo.
a recipe for advanced analytics
The analytics arms race of the past 15 years has taught the
personal auto insurance industry several lessons, but the
lessons have much broader application. In all lines of insurance, the key to sustaining a competitive advantage is to
match price to risk accurately. In this age of analytics, opportunities for using new information sources to improve
