馃攳
Leveraging Real-world and Risk Neutral ESG solutions within Insurance | Numerix Video Blog - YouTube
Channel: numerixanalytics
[14]
Jim Jockle (Host): Hi welcome to Numerix Video
Blog. I'm your host Jim Jockle. Joining
[17]
me today is Alex Marion, VP of Product Management
here at Numerix for insurance products. Alex
[21]
how are you?
Alex Marion (Guest): I'm great.
[22]
Jockle: Alex last time we talked, we talked
about trends in insurance companies and elements
[27]
of onboarding new technologies. And bucking
the trend of overseeing more on the banking
[33]
side and the capital market side, where we're
starting to see a lot more outsourcing and
[37]
investment. In insurance, we're seeing insourcing.
And one of the areas in particular is around
[43]
economic scenario generation. Perhaps, first
of all, could you give us a good definition
[49]
of economic scenario generation? And then
let's talk about some of the trends of why
[53]
insurance companies are starting to inboard
new technologies.
[56]
Marion: Sure. Well there are two dominant
use cases of economic scenario generation
[60]
technology in insurance companies. The first
one, is risk neutral ESG's. Which are risk
[66]
neutral models which are models used for pricing
and hedging, embedded derivatives in insurance
[72]
products such as variable annuities or fixed
index annuities. So, risk neutral dynamics
[76]
are the absolute correct dynamics to use for
valuing these embedded derivatives just like
[80]
they would at banks. The other use case is
what they call real world or realistic scenario
[87]
generation and under these, that have been
around for a while in insurance companies
[91]
and companies are now using a stochastic set
of real world scenarios to project forward
[97]
their liabilities for capital and reserving
requirements. And they're looking to do
[101]
that on an enterprise wide platform so for
things like AG 43 calculations, or risk based
[109]
cap, or economic capital applications, they're
all using these real world ESG solutions to
[115]
project forward their balance sheet across
different lines of business.
[118]
Jockle: So we've also in annuity space have
seen a lot of moving around in terms of people
[124]
entering the business, people exiting the
business, buying different books of business,
[128]
what is a best practice as it relates to building
these books after driven through lack of a
[133]
better term acquisition.
Marion: Sure so there has been a lot of M&A
[137]
activity in the insurance space, a lot of
it has been about rebuilding capital positions
[142]
and trying to decide how do we effectively
deploy capital. So when risk managers in insurance
[149]
executives are looking at the opportunities
in a market they have to have the tools that
[154]
help them turn data information into management
decisions. What lines of business will I best
[161]
deploy my capital to? What other lines of
business should I be de-risking - are they
[165]
consuming capital? And to do that they really
need a consistent enterprise risk management
[171]
framework across different products and different
lines of business and economic scenario generators
[175]
allow them to unify these different legacy
cash flow systems and different valuation
[181]
systems in the single consistent framework.
Jockle: So you've mentioned in essence -- stochastic
[186]
on stochastic issues. Which brings its own
host of challenges in terms of deploying different
[193]
hybrid models and moving away from more traditional
-- black and things of that nature. Tell
[198]
me a little bit more about the onboarding's
of this type of technology and how equipped
[206]
are these companies to maneuver more complex
models.
[209]
Marion: Sure. So if you look at the variable
annuity market or the fixed index market where
[216]
these annuities have these embedded derivatives,
a lot of companies are bringing their hedge
[221]
programs in-house. Right, so they're dynamically
hedging the embedded derivative risk whether
[226]
its guaranteed minimum withdrawal benefits,
other VA guarantees, or some of the fixed
[230]
index annuity crediting strategies. When they
do reserves calculations or hedge strategy
[236]
projection on some of these businesses, they
have to project forward that hedging strategy
[241]
and to do that, they have to have an outer
loop of realistic scenarios to project forward
[246]
the business but they have to revalue those
hedges and the liabilities at each interlope
[252]
node along those real world paths.
So that's the general framework of the nested
[256]
stochastic problem and it can be very computationally
intense. Now if you have a consistent modelling
[263]
dynamic between your risk neutral models and
your real world models, you can build out
[267]
a nested stochastic framework that's consistent,
it's defensible, and it's optimized for
[274]
computational run time. And a lot of companies
now are looking to use cloud computing solutions
[280]
to enhance their nested stochastic capabilities
by overcoming some of the run time issues.
[286]
Jockle: And that's more from an infrastructure
as a server -- burst to cloud?
[289]
Marion: Sure. Right exactly.
Jockle: So one of the questions as a driver
[293]
also the hedging programs themselves and the
strategies being employed are becoming much
[297]
more sophisticated.
Marion: Sure. For example in a variable annuity
[302]
market, a lot of companies have been trying
to move the hedging from the insurer back
[306]
to the policy holder in the form of these
risk managed funds. Whether they are managed
[312]
vol strategies, or capital protection overlays,
they're farming out some of that risk, whether
[319]
it's vega risk, they're moving into the
funds. The funds are de-risking as the market
[324]
goes down as volatility increases and that
translates to a lower greek profile for some
[329]
of the embedded guarantees, and a lower hedging
cost. So, as part of the de-risking theme
[335]
in the variable annuity market, this has been
the central component of that is to move the
[340]
guarantee derivative risk over to the policy
holder in the form of these managed products.
[344]
Jockle: Well, Alex I'd like to thank you
so much for your time today. And of course
[348]
we'd like to talk about the topics that
you'd like to talk about. So if you're
[350]
on twitter, please add us at @nxanalytics,
or on LinkedIn to stay up to date for all
[354]
the conversations that we have, whether it's
our webinar programs or events, or updates
[358]
on our video blogs. So with that, I'm going
to say thank you again Alex.
[361]
Marion: Thank you.
Jockle: And we'll see you next time.
Most Recent Videos:
You can go back to the homepage right here: Homepage