Plenty of Work Remains
Underwriters are being assisted—not replaced—by analytics tools as their jobs
continue to change.
There has always been a fear that humans would be replaced by machines—specifically computers. While entire
industries have been erased thanks to technology, insurance underwriters needn’t worry about their future,
according to a trio of insurance industry analysts. However, underwriters should expect changes in how they go
about their jobs. This month’s participants include Karen Pauli, a principal with Strategy Meets Action; Martina
Conlon, senior vice president in the insurance practice at Novarica; and Karen Monks, an analyst in Celent’s
North American insurance practice.
What Do You Think?
This month’s question:
As analytics tools continue to improve, there seems to be renewed worry that
insurance underwriters may go the way of travel agents. What do you see in the
future for underwriting staffs?
Karen Pauli,
SMA
The underwriting
staff has a bright
future. Roles are
changing, however,
like almost everything in insurance.
Risk is becoming
significantly more
complex. Cyber,
For some time, personal auto un-
derwriting has been fully automated
by many insurers. This has transitioned
underwriters into the role of portfolio
underwriters who manage the un-
derwriting profile of entire books of
business. Not the same role as tradition-
al underwriting, but a challenging and
interesting one.
Technology will support underwriters
as insurers expand their appetite for new
products in a global market. Associating
all of the underwriting risk variables will
be supported by new data sources, but
skilled underwriters will still need to collaborate with customers and distributors
to appropriately manage risk.
Martina Conlon,
Novarica
Predictive analytics
is essentially the
most modern form
of underwriting, a
practice that has its
roots in actuarial tables and predicting
future risk based
on past patterns of
information.
Underwriters will need to become
more efficient and effectively leverage
predictive modelling, however they
won’t be replaced by analytics until
organizations become comfortable
with excluding human intervention and
trusting multiple billions of dollars in
risk pricing to algorithms.
Nonetheless, predictive analytics will
inevitably be a source of pressure, as
predictive scoring is, to some extent, a
replacement for underwriting.
However, it will also be a lifeline. Predictive modelling will afford underwriters a more robust toolset with which
to anticipate risk and forecast product
profitability.
Karen Monks,
Celent
Automation makes
the underwriter’s job
more interesting by
eliminating routine
cases through the
use of standardized
rules to make risk
decisions. Although
algorithms can
analyze the risk of
a policyholder, there will always be cases
that require an underwriter’s expertise.
Underwriting analytics tools review
non-medical information and digital data
such as lab results and build upon past
experience to assign risk. This analysis is
best suited for lower-risk cases.
Since not all cases fit the rules, human
underwriters are needed for these cases.
Human interpretation of medical data
will also remain for the foreseeable
future; the risk for many insurers is considered too great.
What about 10 years from now? The
number of underwriters will decline as
analytics improve, but in my opinion
there will never be a time that human
interpretation of risk won’t be needed.
And don’t forget, underwriters will still
be needed to design, monitor, and revise
the automation rules.