Risk Management Framework in A Life Insurance Company

Download as pdf or txt
Download as pdf or txt
You are on page 1of 18

*

RECORD, Volume 29, No. 3


October Annual Meeting
October 2629, 2003

Session 89OF
Risk Management Framework in a Life Insurance Company
Track: Financial Reporting

Moderator: MAX J. RUDOLPH


Panelists: MAX J. RUDOLPH
ELLEN COOPER
JAMES P. GREATON

Summary: This session introduces attendees to the growing area of enterprise risk
management (ERM). It provides an overview of an ERM framework as well as issues
and challenges of implementation. Regulatory and rating agency reactions are
discussed in depth. Attendees will have an enhanced understanding of what is
meant by ERM in concept and application, various undertakings within the United
States and Canada, issues and challenges to implementation, and regulatory and
rating agency reactions.

MR. MAX RUDOLPH: Welcome to Session 89. Im going to introduce our three
speakers and then well get right into the program.

Im a vice president and actuary with Mutual of Omaha. I graduated from Michigan
Technological University. My focus is primarily on asset/liability management (ALM)
and other balance sheet risks, such as liquidity and capital. I am a past Investment
Section chair and one of the founding members of the Risk Management Task
Force. Ellen Cooper, a graduate of Temple, is a senior manager with Ernst & Young
in Philadelphia, focusing exclusively in the risk and value optimization consulting
area. Her expertise includes corporate risk and corporate investment reviews,
stochastic modeling, variable annuity analysis and stochastic generators. Jim
Greaton is a graduate of Purdue. He is currently Vice President, Annuities Actuarial
_________________________________
*Copyright 2004, Society of Actuaries
Risk ManagementFramework in a Life Insurance Company 2

at Sun Life. Prior to that he was Vice President, Risk Management in charge of Sun
Life's Risk Management function for its U.S. Operations. He has served on the
Academys Committee on Life Insurance Financial Reporting (CLIFR). He was also a
member of the Societys Financial Reporting Section Council and served on the
Academys Equity-Indexed Annuities (EIA) Task Force.

I will focus on the basics on enterprise risk management to set the stage for our
other speakers. A lot of what Im talking about today is based on a book that came
out in summer 2003 called "Enterprise Risk Management: From Incentives to
Controls." James Lam is the author. I recommend it. James is one of the board
members for the Professional Risk Managers' International Association (PRMIA), so
hes very involved in risk management. James has spoken at previous SOA
seminars.

Not everything that Im going to say is from Dr. Lams book, because hes focusing
on how to do enterprise risk management across a wide variety of industries. There
are a few things that Ive tried to expand on, so if you dont agree with something,
its likely something that I have added.

What is risk management? Its an integrated approach to managing a portfolio of


risks, as opposed to a silo approach. Its a framework. Its a process. It's a tool. Its
not an end-all.

While risk mitigation is important, its only part of the story. Enterprise risk
management can be used to optimize returns relative to risk for your product line
or your personal financial situation. It helps you get to the efficient frontier. We
should be using enterprise risk management as a profit center, to suggest solutions
where we can take more risk, make the company more profitable and increase
shareholder value. How can we do things better? We must be aware of the
appropriate risks to measure and manage. Enterprise risk management is driven by
the control cycle. You look at where you are, formulate a plan to make it better and
then monitor the results. You iteratively make it better. Its not going to be a one-
time fix.

You should not always control or reduce risk. Sometimes it makes more sense to
increase risk. If you are ahead of the curve on enterprise risk management, you will
have a competitive advantage in the industry. The key is culture. You need a
common language. You need people talking a common language across different
business units, including the investment department. You need a champion who
walks the walk. If the champion says, "Yes, we need to do enterprise risk
management. But, of course, it doesnt apply to me; it only applies to people below
me," thats not going to work. Employees pick up on those actions. A lot of it is
common sense. The models are important, but you still need to make sure that
they make sense. If an arbitrage-free scenario tells you that the value of a bond
thats going to mature in two days at $1,000 is only worth 2 cents, you should
challenge the model.
Risk ManagementFramework in a Life Insurance Company 3

The alternative to thinking ahead about risk management is crisis management.


You can either wait for the problem to arrive, or you can try to anticipate what the
problem is going to be. Analysis using risk silos is a start, but you need to go
beyond that and look at the interrelationships and correlations between different
risks.

What are the practical reasons for ERM? Its a big part of managements job. If
management doesnt think theyre responsible for managing the risks of their
business, perhaps saying thats the actuary's job, theres a problem. Many
consultants suggest pursuing enterprise risk management because it can reduce
earnings volatility. Im a firm believer that thats only part of the picture. We get
paid, as insurers, to take volatility from individuals. When somebody buys a life
insurance policy from us, that person is reducing his or her volatility and increasing
ours. With the law of large numbers this works out for us. Bottom line, our job is to
buy volatility and get paid for it.

Theoretically, an academic might tell you that an investor can diversify. They would
say that managements job is not to manage all their risks, or to limit some of
those risks, because an individual can always invest in something that has a
negative correlation to that company. Thats true to a point. But you have a lot of
other stakeholders besides the investor. You have all your employees. You have
your policyholders. Its managements job to manage all of a companys risks.

I remember a consultant coming through a couple of years ago and showing us how
great it was going to be if we could reduce earnings volatility by doing something
that they were suggesting. The consultant never mentioned the offsetting cost. You
dont want earnings management. That was really what they were selling at the
time, and a lot of companies have gotten in trouble since the public realized thats
what they were doing. Hedging is good, but reserve manipulation is bad. At GE, as
noted in the Jack Welch books, they could reduce a 15 percent return that
exceeded guidance to 12 percent by increasing their research budget. I dont think
thats good. You lose transparency.

Efforts to maximize shareholder value are driven by culture. If you can build risk
management concepts into your culture, opportunities will be easier to see. People
are going to be more aware of opportunities. Someone may read in the paper that
theres a company that looks undervalued. Maybe they should look at it. Maybe it
goes through and you end up doing a deal that provides an internal hedge. It
provides a disciplined approach and consistency. Thats a good thing. It forces you
to look at risk-adjusted returns. It builds skills internally. If I have a lot of skills in
the corporate area for managing risk, but I dont have any in the product lines, the
company is not better off. We need to have a culture that is aware of risk, where
we are doing the cost-benefit analysis to make sure that we get paid for the risks
that were taking.
Risk ManagementFramework in a Life Insurance Company 4

Let's talk about job and financial security. Senior managers who practice risk
management are less susceptible to external volatility for interest rates or for a big
spike in mortality. Senior managers generally have stock options and invest in their
own companies. If their company goes under, those stock options arent worth
much. So there is a benefit to risk management for senior managers as well. Its in
their best interest. Hopefully, theyll recognize that and do it.

There have been several recent risk management issues for actuaries. Im sure I
am not telling you anything you dont already know. Its important for us not to do
single-scenario, expected value pricing where, if I look 20 years out, it looks great,
but I went bankrupt in the fifth year. If Im in great health 20 years from now, but I
died five years out, who cares? Im still dead. It just doesnt work. The guaranteed
minimum death benefit (GMDB) product feature is covered by other sessions at this
conference. You need to keep track of whats going on, if youre a U.S. actuary,
with the American Academy of Actuaries RBC C-3 Phase II project. Its expected to
move forward soon. Then it will be up to us, the people in this room, to implement
it. There will be a lot going on next year with that.

When we first started writing long-term care, who knew that product was lapse
supported? Hopefully, nobody who was pricing it. There are also bank annuity-type
products, primarily deferred annuities, where weve given out options but havent
charged the right price for them. One of my mantras is "Never assume the
policyholder is stupid." Even if they act that way today, they may not act that way
tomorrow. Were seeing some experience in our deferred annuity block right now
where its in their best interest to stay due to the low interest rate environment. We
had a floor in our lapse curve, and were going to take it off. Its no longer
appropriate. They figured it out. The same thing happened with mortgage-backed
securities. I occasionally give talks to our financial staff, like the accountants and
investment folks. I'll ask who has refinanced their home mortgage in the last couple
of years? As the ALM guy at an insurance company, I hate those people almost
everyone with a home mortgage. Thats the risk. Think about how easy it is to get
the information that it is in our best interest to refinance. Its in USA Today. Its
even on the front page of the local weekly paper of villages with 5,000 people. Its
everywhere.

I want to focus on what James Lam was talking about. He discusses seven basics of
enterprise risk management for senior managers. I think he did a pretty good job of
coming up with them.

1. You want to know the business that youre in.

2. Make sure there are checks and balances.

3. You want limits and boundarieswhen should you stop writing variable annuity
business with GMDB dollar for dollar?
Risk ManagementFramework in a Life Insurance Company 5

4. Keep your eye on the cash. (Lam included this quote, "Cash is king. Accounting
is opinion." I love it.)

5. ERM provides a yardstick. As we view it more and get better at it, well
understand which pieces of risk management are leading indicators and which ones
are lagging indicators.

6. We want to incent the desired performance; we must put incentives in place that
encourage people to make the right decisions. I know, as actuaries, we struggle
with that a lot. One of the things I liked best about this book is that, although I
didnt learn a lot by reading it, it gave me lots of quotes that I could go and say,
"See, its not just me. Here, look at this guy. He wrote a book, they published it
and its saying the same thing." Thats very powerful with some senior managers.
They dont want to hear it from their actuary, but they will listen if somebody has
been published.

7. Finally, there is the distinction between hard and soft risk management. Soft risk
management includes enablers. What types of things enable you to do risk
management? They are things like infrastructure, independent risk functions,
oversight committees, audits and risk assessments, procedures and policies,
systems and models, with reports ready to go on a regular basis. My staff doesnt
like it when I say this, but we need to get to the point where we can push a button
to produce our reports. Push a button and it will give us the output overnight. Then
we can spend our time analyzing it. There are some companies that are at that
point. Its very impressive. Theyre able to spend their time on the analysis. If we
can spend our time on the analysis instead of on the data collection, think what we
could accomplish. It improves our timeliness. It improves our effectiveness.
Enablers also allow you to limit the exception process. Say were only going to sell
life policies up to age 80, except for friends of the chairman. You cant allow that
and have it work.

Then you have your drivers, the hard form of risk management. You need a
champion, and it needs to be somebody at the senior level. You need that person to
recruit other champions. You need champions throughout the company. You need
the commitment, because that is whats going to drive enterprise risk management.
Culture and values are important. Common sense is core. Design incentives so that
it drives that same common sense process.

You want to facilitate open communication. One question that I hear a lot in the
industry is, to whom does the chief risk officer (CRO) report? The chief risk officer
might report to the chief financial officer or the chief operating officer. The CRO
really needs to report to the board, or report to the CEO and have direct access to
the board. If you see something bad happening, you need to be able, with
impunity, to go to the board and tell them that there are concerns. One of the
drivers, and I know I am preaching to the choir, is learning. Be open to continuing
education. The ERM field is changing quickly. Emerging issues include fair value
Risk ManagementFramework in a Life Insurance Company 6

reporting, C-3 Phase II, stochastic morbidity and stochastic mortality. Five years
from now, todays skill sets wont be good enough, even for the people who are
here. We have to continue to learn, including the people who are experts today.
Well have to continue to expand. Again, we want to reinforce it all with incentives.

This is a quote from the Lam book that summarizes the benefits of risk
management. "Companies that understand the risk/return economics of the
business can take more of the profitable risks that make sense for the company and
less of the ones that dont." The part I add, "and know which are which," is
important too.

What would I suggest that you take away from this? If you consider yourself a risk
professional, put together a list of what keeps you up at night. Its interesting that
the rating agencies have picked up on that. I always get to go last, not surprisingly.
I discuss ALM-type issues. Often theyre late for a flight, theyve been running long
and they need me to go very quickly. So theyll stop and say, "We dont have time
to go through your talk, but what keeps you up at night?" If you have a "top 10"
list, a risk scorecard or something that talks about emerging issues, things that
youre reading about in the industry and affects your company, thats all very
useful. Theres a lot of value to that.

Dont wait for modeling perfection. Thats always been the case, going back to
before cash flow testing came around. You can work on your models trying to
recreate your day-by-day cash flows, but youre never going to get there. Do the
best you can today, and then figure out ways to iteratively improve as you go
along, whether its next quarter, next year or whenever. Theres always room for
improvement.

MS. ELLEN COOPER: Im going to talk about risk management in the context of
the life insurer. Youll see that some of the themes Ill be discussing are going to
overlap with some of the ideas presented by Max and Jim. When we think about
attempting to solve the problem of measuring and managing risk, many of the
processes and conceptual ideas about it are really quite similar. Its in some of the
details where we get bogged down.

The first thing that we need is a mission statement. What are we trying to achieve
when we talk about risk management? What are our goals? By a show of hands,
how many of you within your organizations have some kind of an enterprise risk
management framework or process on a regular basis? Good. Id say thats
probably about half of you. For those of you that do, and also for those of you
striving to have some kind of a framework, I recommend thinking about what the
goal is. What is our mission? What is it that we want to achieve? While youre
defining the mission, you want to think about having consistency across the
organization. What is my preferred measure of risk, and is it consistent with the
people in the U.K.? Is it consistent between the life and annuity lines?
Risk ManagementFramework in a Life Insurance Company 7

I want to make sure to aggregate my exposures across lines consistently, so that I


can appropriately assess my diversification and concentration. Finally, I want to be
able to define my global risk tolerance framework across the organization. I need a
consistent definition, so that when I look across the organization globally and
enterprise-wide, or by line of business, its in a consistent framework.

When deciding on my mission, I also need to decide what kinds of information


senior management and my unit need from me. What will it enable them to do? If I
just print some glossy report on a quarterly basis, give it to somebody and nothing
ever happens with it, then its not adding much value. In this particular framework,
we have the ability to add a lot of value if were able to provide the right
information, if its explainable and accessible.

In addition to having a mission and understanding what our long-term objectives


are, we need to know our definition of risk. As an organization, we need to be
consistent about what risk is. Is risk looking at tail scenarios? Is it the 1 percent
probability of going insolvent? Or is it the 10 percent chance of negative internal
rate of return (IRR) under a new product? Is it what happens if my earnings are
floating around too much? Or is it something else? Understand as an organization
what its risk definition is. I cant manage it if I dont know what it is.

I have to look at all the different risks that we need to be managing in our
enterprise. This includes all types of insurance risk, credit risk, investment risk,
business risk and operational risk. We need to do it all simultaneously.

Once I've defined my mission and my risk definition, and Ive identified what all my
risks are across my organization, simultaneously, I also need to define what my
organizational structure is going to be. We need to have clear roles. Max talked
about that. We need to have good people. We need to have strong leadership. We
need to understand our decision process regarding taking new risks. How are we
going to examine risk issues? What is the process? How is our analysis going to
drive management actions?

We need somebody to be the champion, as Max said. We need a CRO, a head of


ALMsomebody who can be the overall enterprise risk management leader. By a
show of hands, how many in your organizations today have somebody who is either
performing the role of CRO or some kind of ALM leadership position? About a third
of you. That is consistent with other analysis that weve seen.

Im going to walk through an organizational structure. Its intended to be one


example. Its an example of how you can put things together. The CRO should be
responsible for leadership, and then should be able to delegate execution to the
operating divisions. Somebody should be in charge of designing investment
strategy and product pricing, and for execution of modeling, measuring and
monitoring the risks.
Risk ManagementFramework in a Life Insurance Company 8

In this organizational structure, we have included a board of directors. They give


the responsibility of day-to-day risk management to senior management. Weve
seen examples in organizations where they have a senior executive committee
focused on risk management. They appoint a CRO, or some champion, who is head
of the organization. They also make sure that the risk management infrastructure is
in place. Its the CRO's job to make sure that the process is in place. Ill talk later
about how that means having a written policy on risk management, understanding
how to set controls and how to have appropriate tools and processes around them.

The CRO needs to make sure that there is a consistent framework across the
organization, including measures and assumptions. The CRO needs to understand
where models are different, why, and where its material. They need to know how
to aggregate. If I have a PTS model in one part of my organization, a MoSes model
somewhere else and an old Excel spreadsheet that somebody wrote 14 years ago
somewhere else, I have to figure out how to aggregate all that and understand
where the differences and inconsistencies are. If Im doing stochastic modeling, I
have to make sure that the generation of economic scenarios is consistent or, if its
not, I need to understand what the differences are. Are my interest rates parlayed
with equity risk? The CRO is responsible for setting those policies in place upfront.

The CRO is also responsible for assessing risk exposure relative to risk tolerances,
monitoring them on a regular basis and then developing a routine process to
evaluate and control.

As I said before, the operating divisions are then responsible for the execution.
They make sure that the real work is actually happening on a day-to-day basis, and
happening efficiently.

We talked about defining risks and we talked about looking at all our various types
of risk. We also need to create an inventory of the risks. We need to make sure that
we have included everything. One way to do this is to qualitatively rank the risks.
Categorize them and prioritize them, one through 10 or one through five. What are
the most important risks? What are the things that keep me up at night? What are
the things that I worry about most and that I need to manage most carefully?

Then I have to figure out which ones are easy to quantify. We have solved how we
can quantify interest rate and equity risk, but others are more difficult, like
terrorism, operational risk, credit risk for longer than a one-year time horizon, and
trends in mortality. I can put those off on the side and deal with them as more
difficult, long-term situations. In the meantime, I can make sure that, for those
that I can solve, I begin now. Using my prioritization, I start to fix the problems
that are high on my list, monitoring and measuring them now.

So far, I have identified my risks and decided which ones Im going to quantify
now, later, or never. Now I need to understand what I want to do with the risk. Do
I want to take it and actively manage it? Do I not want to take it, and by not taking
Risk ManagementFramework in a Life Insurance Company 9

it, do I want to make an active choice not to take it? I might want to make an
active choice, for example, to avoid any living benefits on my books or not to sell
GMDBs anymore. Or do I want to take it and transfer the risk? Do I want to have
some kind of a hedging strategy, some kind of reinsurance or other mitigation
strategy?

Then finally, there has to be some kind of process to re-evaluate the initial
decisions on a periodic basis. Do they still hold? Do I still want to hedge? Is it too
expensive for me to hedge? Do I want to get out of the business? Do I want to
accept the risks?

As we said before, the CRO is responsible for a written policy and developing
standardized reports. It's very important to make sure that the information being
communicated is understandable and readable to management. Investment
guidelines and strategy could be done by a CRO or by the chief investment officer,
depending on the organizational structure. The CRO would also be responsible for
processes around controls.

In terms of measurement, I need to make sure that I have good models. Again, the
CRO is responsible for making sure that models are consistent across the
organization and to be aware of any differences. Im going to perform risk
measurement, and Im going to do it over multiple time horizons. Im going to look
at things like interest rate, equity, credit, foreign currency (if I have it), mortality,
morbidity, and policyholder behavior. I'll make sure to do appropriate stress tests,
sensitivity tests, and back tests so that I really understand the risks.

I also need to decide what kind of metrics I want to use in my risk measurement. Is
this going to be an embedded value (EV) framework? Embedded value-at-risk
(EvaR)? Risk-adjusted return on capital (RAROC)? Im going to make sure that I
report on a regular basis and that I review my exposures and tolerances on a
regular basis.

I want to make sure that I have appropriate controls. This is not intended to be a
plug for consultants, but there should be an independent review of our models.
There should be some other person outside of the regular day-to-day function,
whether its somebody else in the organization or an external consultant, who
reviews the models, critiques the process from time to time and gives insight into
places where there is room for improvement. There will always be room for
improvement. I need to make sure that my models are validated. If I have
something in my model that doesnt make sense, obviously I want to make sure
that I fix it before I start doing all my risk analysis.

Finally, I want to make sure that my policies and guidelines are clear and that
people can understand them. Somebody other than the person who has written
them is going to read and interpret them. If theyre not extremely clear, they could
be misunderstood or gamed. I want to make sure to have exposure targets around
Risk ManagementFramework in a Life Insurance Company 10

asset classes, credit quality, duration and convexity mismatch, and so forth. This
way, when Im measuring my exposures and tolerances, I know when Im out of
balance and need to take some kind of action.

I want to make sure that reporting is timely, automated and accurate. I can
develop routine reports that are not ad hoc. Its enough that we have these big
complicated models that can take 24 days to run. Once Im done with that, I then
need to put everything into some kind of report. That process should be automated
and standardized to the extent possible.

MR. JAMES GREATON: I currently work in the U.S. Annuities Division at Sun Life.
Before that, I was vice president of Risk Management. Im going to look at how Sun
Life is trying to implement some of the ideas that Max and Ellen were talking about.

I'll give a brief background on Sun Life. Its a Canadian insurer. Therefore, its
subject to regulation by the Office of the Superintendent of Financial Institutions
(OSFI), the regulatory body in Canada. We have three different types of entities
that we have to worry about in the United States. We have a U.S. branch of a
Canadian company. We have a U.S. subsidiary of that company. We have two other
subsidiariesKeyport, which is a life insurance company, and IFMG, which is a
marketing organizationthat were recently purchased by Sun Life.

Its different up north of the border. OSFI is a federal regulatory body that
regulates all financial institutions, not just insurers but also banks, brokerage firms
and the like. Their exams include an assessment of the risk management
capabilities of the organization. They look at the inherent risks and the risk
mitigants that the insurer has put in place. Theyre currently in the process of
implementing a risk management rating system, so theyre going to rate each
financial institution, not just insurers, as to how they manage risk. So theres a real
imperative for us to have a risk management process up and running. Its
something the regulator has mandated that were going to do, and OSFI is going to
rate us on how well were doing it.

As Ellen described, its important to have a framework. Its what our approach to
risk management is going to be and, more importantly, who is going to be
responsible for what and how that is going to be delegated down into the
organization. I want to stress that this is how Sun Life has done it; its not the only
way. You could come up with a different framework and you could come up with
different responsibilities. The key is that you have a framework and assign
responsibilities to individuals.

We have written policy statements that we call "consolidated risk management


policies" (CRMPs). The board approves them, and the board gives specific authority
to management to implement policy statements and risk categorizations within the
written policies. Once the board has promulgated a risk management policy, the
central risk office will then try to slate it into different risk categorizations and come
Risk ManagementFramework in a Life Insurance Company 11

up with policy statements. We have a grid that we put together that uses red,
yellow and green lights. Red are those risks that we want to avoid. Green are those
risks that are okay to take. Yellow are the risks that you need senior management
approval to accept. Then weve put together a process.

The policy statements will state what kinds of risk we take. What kind of risk is it,
and does it fit smoothly into the current categories that we have? Whats our
tolerance level for it? Are we going to accept it or not, and how much will we take?
It also defines a maximum tolerable loss in terms of volatility of annual earnings.
How much pain are we willing to take in any one year for having the risk on board?

Ill step through what our high-level categories of risks are under our consolidated
risk management policies. There's capital adequacy risk. Do we have sufficient
capital to absorb expected fluctuations in the business? We have asset risks, such
as credit or liquidity. We have liability side risks, such as mortality, morbidity,
lapse, and expense. We have asset/liability risk. Do we invest properly to match the
cash flows of the liabilities? Then we have operational risks, such as fraud, market
conduct, legal liability, terrorism, disaster, HR concerns and strategic concerns.

I thought Id take a moment to review so you can get a feel for the framework and
how its designed to operate. Lets take credit risk. We have a consolidated risk
management policy about investment risk. In it, there will be a maxim, or
statement, that we take on investment risk in order to minimize the costs of our
liabilities. In order to maximize our return on investments, we need to invest in
obligations of corporate companies. They have credit risk associated with them.
Then well assign responsibility to a particular position, not to an individual, in this
case the chief investment officer, to investigate credit risk and generate maximum
allowables for it. Then theres a policy process in which the chief investment officer
will come up with, for credit risk, tolerances within certain investment portfolios for
how many names are allowed or what exposure to particular names is allowed.
There are very specific limits defined. Then theres a process to monitor whether
were complying with the limits. Its a top-down approach. At the top, authorization
is given to an individual to set the risk parameters. Then the parameters are
defined, followed by a process to see if were complying with the parameters.

In a general risk management framework, the first thing you need to do is to figure
out what risks you have. You then need to select a risk management strategy. You
need to implement that strategy, report on it and monitor it. Then you start the
cycle again. You then have to re-identify or reassess the risks, select the strategy
and go around again.

I want to stress what Ellen said, that identifying the risk is not a risk management
function. Risk management is the leadership in the organization that puts the
process in place. It has a public presence to enforce the culture that you're trying to
impose on the organization. But its the organization itself that has to take
ownership for the risk management process. Its not a risk management function to
Risk ManagementFramework in a Life Insurance Company 12

identify the risks. Youre providing a framework. You make sure that the risks are
identified and tracked. Its the other folks on the front lines that know the business,
see the stuff coming, and go with it.

The real intent is to have a strategic business unit (SBU) own the risk, not the risk
management area. If risk management owns it, then no one is managing it. Other
areas can help us identify our risks in addition to the strategic business unit. That
could be audit, compliance or controllers. Investment folks or sales folks can also
contribute ideas.

We have several different processes in place at Sun Life that help us identify risks.
One is a "top 10" process. Annually, each strategic business unit and each
corporate line function, such as the finance area, the compliance area or the audit
area, will identify what they think the top 10 risks are that their organization faces.

Our product design process is also a place to identify risks. It's where we price a
product and design it, but part of that process is to identify the risks associated
with a new product. Our compliance area also identifies risks. They notice a lot of
the risks because someone else in the industry has messed up. Generally there is a
regulatory body that is either put in place to monitor that risk exposure or
regulations have been put in place to avoid it. So compliance is a useful function for
identifying risks and having them appear on the radar screen.

Weve also instituted a series of regular reports that helps us identify when risks
get out of kilter, and that can spur identification of risks. We produce maximum risk
tolerance limit (MRTL) reports, earnings at risk (EaR) reports, and an annual
dynamic capital adequacy test (DCAT), which is disaster scenario testing.

In Canada, they also have something called the SSBFPs, the Sound Standards of
Business and Financial Practices. Its a questionnaire put out by OSFI that identifies
if you have various management practices in place. That can be a good process to
identify risks. Either theres not a management practice in place to handle
something on OSFIs list, or a company response says that the process was in place
but failed in the last year because something else popped up.

We also do an annual assessment of embedded options in our organization, where


we look at the liability side of the balance sheet and assess each product for its
embedded option features. Once youve identified a risk, management has certain
options. Ellen went through these as well. You can avoid the risk, manage the risk
with a mitigation strategy, transfer the risk, accept it, or try to eliminate it all
together. You can eliminate it by fully transferring it or avoiding it.

You cant have risk management without a management action plan. Within our
consolidated risk management policies, there is authorization for management to
deal with risks. The risk management policies call for specific action plans, with
clear lines of accountability for each risk. In other words, if the risk is identified in a
Risk ManagementFramework in a Life Insurance Company 13

"top 10" process, through an audit or an SSBFP, we need to identify the risk and
have a written paper trail of what the action plan is. Are we going to avoid the risk?
Are we going to transfer the risk? Who is responsible for implementing the action
plan? It is no good unless its executed, so these action plans are usually put on
peoples performance objectives.

Part of the process is to have regular reporting. We have a series of committees


within Sun Life that meet regularly and receive regular reports. Theres a senior
leadership team as well as risk management committees. Ultimately it all rolls up
into a report that is given to the board once a quarter.

Part of the process is actual incidents, things that have arisen during the quarter.
There are current assessments. Each actionable item will be assessed as to whether
anything has been done on it. The person who is held accountable reports to the
risk management committees on that item. Of course, there are status updates on
the action plans.

I said there are risk management committees; we have more than one. We have a
U.S. Risk Management Committee. We have a Central Risk Steering Committee,
which is made up of the risk management officer in each one of the national offices,
along with the chief risk officer for the company. We have an Executive Risk
Committee, which is comprised of the chairman, president and chief financial officer
of the organization. The chief risk officer reports to the Executive Risk Committee
once a quarter. The Audit Committee of the U.S. boards gets a report once a
quarter, as does the Risk Review Committee of the parent board. Of course,
Internal Audit also does periodic reviews. Part of their review process is to review
the "top 10" process, but they will also identify separate initiatives each year, most
of which have some risk piece to them.

All this corporate governance does not come at zero cost. Theres a staff. We have
a corporate staff that sets policies within the consolidated risk management
policies. There is a vice president in charge of risk for each of the national
operations in the United States, Canada, U.K., and Asia. They monitor compliance
within the national organizations.

Cant we avoid all risk? No. As Ellen and Max have alluded to, acceptance of risk is
an essential part of our corporate strategy. The goal is to earn a competitive return
at acceptable risk levels. Effective risk management protects and enhances our
value. It helps us meet that return goal within acceptable risk levels.

There are some practical pitfalls that should be avoided. One is if there is a risk
identified, but its not "owned" by someone in the business unit. If it is a corporate-
identified risk, nothing will get done until the business takes ownership of the risk.
In order for risk management to be effective, it has to be embedded in the culture
of the organization. It has to be something that is bought into by the organization
and by the strategic business units. They have to see that it adds value to the
Risk ManagementFramework in a Life Insurance Company 14

organization.

Management at the senior-most level must support risk management. They must
look for the reports, read them, and give you feedback. They have to make sure
that if a risk is identified, a strategy is identified and accountability assigned, that
there are incentives for the individuals in the organization to follow through on
those plans. If there arent incentives, it wont be done.

The framework must be understood. There have been times when portions of our
framework have been confusing or contradictory. That doesnt help actuaries, in
particular, when youre trying to come up with a product development process and
you have to comply with two different standards, which dont jive with each other
internally.

As actuaries, we tend to think of risk in terms of things that can be quantified. We


like to manage asset-side risk because we can come up with distributions of
investment returns or interest rates or credit losses. A lot of risks in organizations
reside outside of typical actuarial issues that we tend to study and develop
probability distributions for. The operational-side risks, the risks of a sales
organization, and the risk from competition invading your particular market niche
are all types of risks that you need to have on your radar screen and that need to
be part of this process. Dont be afraid of a risk because you cant quantify it. You
can probably quantify it within big boundaries, such as its less probable than a
tornado. Thats fine. Put it on the list and assign responsibility for someone to be
thinking about it. Have an action plan in case it comes to fruition, or be thinking
about how you can avoid it. Thats a real helpful thing to have in your organization.

MR. SHELDON LAVEMAN: How much should an insurance organization spend or


budget to build one of these frameworks? How much might we spend on an
ongoing basis? I think that Sun Life is a good example, because its multijurisdiction
and multiline. How long might it take until you have this framework in production?
As much as there was a comment that we shouldnt be fearful of something we
cant measure, how can you actually measure whether this thing is working? Is
there any research that demonstrates its actually effective at all?

MR. RUDOLPH: Ill answer the last one first. I think there are plenty of examples
where if you dont have it, bad things have happened. Enron was thrown out as the
example of a company with a chief risk officer. It comes back to the culture and the
champion. Did anybody listen to the chief risk officer?

MR. GREATON: How much do you want to spend? You probably already have a lot
of the quantitative tools in place. You probably already have tools that manage and
look at ALM risk, credit risk, or underwriting risk. Its not that expensive of a
function to put in place. You do need some staff. You do need a chief risk officer. If
youre involved in multiple jurisdictions, you probably need a person in each one.
But that persons primary job, at least to get off the ground, is merely to collect the
Risk ManagementFramework in a Life Insurance Company 15

information thats probably already embedded in the organization and to surface


that information to the top. It takes some time to put that framework together.
Some work has already been done in books like the one Max mentioned. Theres
time spent studying, putting together some documents, running it by senior
management, and getting the board to approve it. Its not going to happen
overnight. Its going to take you a year or two. Your first cut at it is going to be
wrong. Youre going to redo it. But its doable.

MS. COOPER: Ill add that if you have nothing in place, the best way to start is by
tackling one line of business and deal with some of the process-type questions. I
think if you did it in the right way and it was a proactive effort, you could have
something up and running in several months for one line and also have a first
attempt at a policy. Within probably three months, you could be looking at
stochastic modeling results to measure things like interest rate and equity risk for
one line of business. Then you add from there. It also allows the company, your
organization, to buy in to whats happening. Once they realize the power of what
youre doing, they will want to see it across the whole organization. They will be
hungry for more once they see whats happening.

As a follow-up to Maxs comment, I think if you dont measure it, you cant manage
it. Without being able to know what the number is, we have no way to know
whether or not its good, bad, or indifferent.

MR. RUDOLPH: You want to leverage off of the models you already have. If youre
doing cash flow testing already, a lot of your financial models have already been
created. Where the struggle comes in is how to do the operational risk and how to
correlate that back to the financial risk. In terms of cost, we have to convince our
senior management that its not really a cost because the return is going to be
greater than the cost to the unit.

MR. LAVEMAN: Can one of you give an example of a framework in place that led
to the successful avoidance of a risk and therefore, you did achieve a benefit that
exceeded the cost?

MR. RUDOLPH: Several years ago, when Long-Term Capital Management was
going down and interest rates were dropping, we looked at whether we should buy
some floors to back our deferred annuities. We said the volatility cost was too high
right then and we would be better off going naked on that. So we did. We saved
quite a bit of money by not doing that.

MR. GREATON: We did the opposite. We bought a series of hedges to hedge our
GMDB costs on the variable annuities because it violated our maximum risk
tolerance limits for earnings. Theyre currently in the money. We havent sold them
yet.

MS. COOPER: Companies often use stochastic modeling to do things like manage
Risk ManagementFramework in a Life Insurance Company 16

interest rate risk. We help companies on a quarterly basis look at things like the
duration and convexity of their portfolios. They actually make decisions about
investment strategy, new products, and interest rate crediting strategies based on
being able to duration match. Thats part of the risk management process. On the
equity side, how much equity risk exposure do I want compared to my interest rate
risk? We do a lot of that kind of work. We help companies figure out what
percentage of risk they actually want to take, which is really an active decision.

MR. RUDOLPH: There is a term that we havent used today. "Risk appetite" is a
concept that is becoming very popular. It means choosing the level of risk that you
want to have, as opposed to letting it choose you.

MR. STUART WASON: Could the panelists give their reflections on whether its
better to have a separation between the role of chief actuary and the role of the
CRO?

MR. GREATON: Certainly at Sun Life, I think its been beneficial to have someone
who is primarily focused on risk as opposed to the role of the chief actuary. The
chief actuary has enough on his plate dealing with valuation standards across
multiple countries, as well as looking at pricing standards and the like. There are
many areas of risk that are not in the traditional actuarial domain. You dont need
an actuary to be a chief risk officer. However, its a good opportunity for an actuary
to step out of a traditional role and broaden his or her experience. Thats been a
good thing for the person that has stepped into that role. So I think its a positive
thing, and these two people certainly have enough on their plates to keep them
occupied.

MR. RUDOLPH: I personally think that in an insurance company, the chief actuary
is a very good candidate for chief risk officer because of the skill set that we have.
So much of the dollar risk is financial, and the actuary likely has a better feel for
that than the accountant or even the investment officer. But I dont think it should
always go to an actuary.

MR. SEAMUS CREEDON: I have a comment and perhaps a question for Jim. Our
equivalent of OSFI in the U.K., our integrated regulator, is the Financial Services
Authority (FSA). As an aside, they publish reports on insurer risk management
practice that are worth reviewing from their Web site. They regularly draw contrasts
between banking industry practice, which they generally like, and insurance
industry practice, of which theyre generally critical. They make the point that the
best banks, particularly the investment banks, are ahead in terms of developing
risk management frameworks, whereas the insurance industry generally is in a
position of trying to guess what will satisfy the regulator. Thats the comment
leading up to the question. How much of Sun Lifes approach was driven by OSFI
versus Sun Life running ahead and creating good practice?

MR. GREATON: I think its a combination of both. OSFI chastised Sun Life for its
Risk ManagementFramework in a Life Insurance Company 17

risk management process a few years ago, and that spurred the process. They
werent too helpful in directing us as to where we ought to go, so it is an internally
developed process that we came up with. Its something that they are a lot more
satisfied with.

MR. JOHN SCHREINER: In discussions Ive had with the rating agencies, the
enterprise risk management processes that theyve seen at insurance companies
have underwhelmed them. What feedback have you gotten from the rating
agencies?

MR. GREATON: Theyve been positive with what we've put in place. It may be
something of a comparison to other insurerswe have something. Theyre
impressed that we quarterly measure our maximum risk tolerance and earnings at
risk. These are things that are not done typically in U.S. insurance companies. So
weve gotten some positive strokes from them, but it may be just because were a
little ahead of the rest of the field.

MR. RUDOLPH: What we find is that we get some very nice comments in the
rating agency write-ups. Over time, what weve tried to do is put together a
consistent presentation so they see the same thing from year to year. Many of
them are learning at the same time as we are. Im finding that were getting better
questions every year. Also, because we show them things and ask them every time
if there is something theyd like us to do, as they start going through and creating,
for example, new capital tests this year, weve gotten on the list and gotten to see
those ahead of time. We've had some comments on them and have been allowed to
comment on them. We dont always get what we think makes sense, but they
listen. If they dont agree with us, they seem to look for verifying opinions from an
unbiased source. I think theyre very open to companies that are doing this type of
work, and I think it is reflected somewhat in the rating agency reports that we get.

MR. MARK ABBOTT: There has been a lot of volatility in the market, especially in
the last several years. In your board discussions defining risk appetite, how do you
actually communicate that, what do you think is the best way to communicate that,
and what role would scenario analysis have in that process?

MR. GREATON: Once again, what we have done in our maximum risk tolerance is
to communicate it purely in terms of how its going to impact Canadian GAAP
reported results. In other words, a certain shock to the equity markets or interest
rate markets is going to impact your Canadian GAAP earnings by X million dollars.
Thats what has been communicated up directly to the board level. Theyve come
back with directions that we should not allow it to vary more than X, and thats
been translated into action.

MR. RUDOLPH: Sun Life is quite a bit ahead of where we are. We frequently
struggle to get our message to the board. There's a book coming out called
"Asset/Liability Management of Financial Institutions" by Leo Tilman at Bear Stearns
Risk ManagementFramework in a Life Insurance Company 18

(available through Institutional Investor). A number of actuaries wrote chapters in


it. I think it will be a good read. Its designed to go beyond insurance companies
and talk about banks. I think it will be a good reference point for us to look at
where were at, relative to other financial institutions.

MR. ABBOTT: My personal view is that there needs to be a balance in risk


management between the various competing risk management perspectives. I
think the chief risk officer has to be someone who is very inclusive in their thinking,
to encompass not only the various financial reporting needs, what the regulators
want to see and what the rating agencies want to see, but, ultimately, what is
going to make the company perform best in the long run. The CRO must have the
long-run perspective in addition to the short-run perspective. There needs to be a
balance of economic valuation perspectives and these accounting issues. I want to
get your opinion about these two competing interests. How can you best mitigate
the risk to the organization of doing something incorrect that will damage it in the
long run, while balancing these competing interests in the short run?

MS. COOPER: Thats part of the challenge, Mark, no question. Some of what weve
helped clients with is looking at the risk under different frameworks. We're looking
at short-term, long-term, GAAP versus stat versus economic value. Companies are
struggling with the question, if Im measuring economic value by calculating a fair
value of my liabilities and Im looking at market value of assets, can I manage to
that? Its not really how the world works right now. I cant really manage to an
economic value. I can manage to my GAAP income, but economic is more difficult
at this point in time. I think it is part of the challenge and the balance. Youre
rightits very complex. There are lots of different issues, lots of competing needs,
short-term versus long-term, different accounting relevancies. Cash versus
accounting is a whole other issue. Just having these frameworks in place, you can
really look at all the different types of analysis.

MR. GREATON: Part of what makes the job fun is that you have the different
frameworks. My old boss didnt like to have multiple models because he didnt want
to have multiple answers. I think getting the multiple answers sometimes helps you
understand the problem better, because you come at it from several different
points of view, and thats helpful. That stimulates the discussion and helps you
understand your business better. If youre going to narrow in on a single constraint,
then youre going to miss a little bit of the richness of the analysis.

You might also like