BY ERIKA MORPHY

Thomas Walsh: As we've gone through what has been a very difficult credit cycle, we've been enduring low interest rates. It has been particularly challenging on the earnings front. I was wondering, relative to those issues, how you felt the life sector's asset/liability management skills have fared over the recent few quarters. Have they improved? Have they been disappointing? And second, are there any systems or techniques that you're seeing out there right now that you consider state of the art, that maybe we should be asking insurers about?

Julie Burke: Our perspective is that overall capabilities have been improving and that most people would point to things like modeling robustness and the functionality of computers - the more powerful hardware and more complex software as well as the ability to run more and more scenarios in a short period of time. So there's the technology side.

But while all that is true, we would suggest that the biggest improvement in ALM (asset/liability management) stems from the consolidation of the industry and the fact that more and more of the business is concentrated in larger, more diversified companies that, frankly, have better capabilities.

What we try to understand is how well a company can measure risk and how well, once measured, they can manage the risk. And certainly, in terms of measuring, we look at the modeling that they do. In terms of managing, what we're talking about is really how the business is brought onto the books, how it is hedged, and how it is kept on the books over time.

In our view, the best hedge is product diversity. But we do see companies using some derivative products much more so than in the past, although it is not very prevalent. Derivatives are still primarily used as an overlay to the existing asset portfolio to address duration mismatch, to swap out fixed and floating rates, or to address currency risks. We haven't really seen a lot of derivative products being used to protect against some of the extreme interest rate scenarios.

We would say most of the larger companies had very good internal capabilities and that they, as well as the rest of the industry, tend to rely on third party software to measure and manage ALM. I think the industry leader is Tillinghast, which I think has about a 75 percent market share. So that's certainly the product that most companies use, at least for some portion of the ALM interest rate management as well as projections.

The most recent technological advances have come in the area of stochastic testing - in other words random scenario generation as opposed to some of the more traditional….what we call deterministic scenarios whereby they're very specified interest rate scenarios. So, certainly, some capabilities have been improving on the technology side as well as from a diversity perspective with the concentration of business into larger, more diversified players.

Thomas Walsh: One of the key advantages that rating agencies have is that you are "above the law" speaking with these companies and obviously go into detail with respect to their portfolios and ALM processes. So the question this brings to mind is: As an insider, what are some of the key variables that you're considering when you rate an insurance company and evaluate their ALM processes and skills?

Julie Burke: When we typically evaluate interest rate risk, we do it in the context of the insurer's overall asset liability management. Therefore, we look not only at the breakdown of asset and liabilities but also at the distribution channels and the policyholder base. We do this because how the product is sold, and who owns the product, can really effect product behavior. We also evaluate the potential for both disintermediation and spread compression under various interest rate environments.

We consider both duration matching risk as well as liquidity risk. We feel we need to evaluate both near term and intermediate term liquidity sources and uses and to do it in under-stressed scenarios. And when looking at liquidity sources, we're looking at both on-balance sheet liquidity as well as backup liquidity - things like lines of credit, rep arrangements, affiliate support.

And also we look at liquidity elsewhere in the organization, or alternatively, liquidity calls. For instance, does the parent company have debt refinancing issues? Are there money-losing affiliates within the structure that need to be funded?

In our view, the best place to start an analysis of a life insurer's ALM risk is really with the product portfolio - and that's both the in-force business as well as what the company is writing today. I think only when you understand the product and the optionality within those products can you truly understand the risks. Now certainly, and it almost goes without saying, understanding and knowing the investment side of it is very important. Our sense is that this side of the equation has gotten and typically gets the proper attention. And quite frankly investment analysts - whether it be fixed income analysts or equity analysts - are typically more comfortable analyzing asset risks than they are liability risks. I think they tend to leave liability risks to the actuaries.

In our view, the best source of information for this are the actuarial memorandums that Michael Barry talked about (see accompanying box entitled "Seven Interest Rate Scenarios below). The best part is their very detailed descriptions of the products. Once the risk levels are determined, what we want to try and understand is how the risk is managed both in terms of the capabilities and the strategies. So, in other words, ability as well as willingness to manage risk.

Insurance regulations took an early step in their development regarding asset/liability management, according to Michael Barry of Fitch Ratings, when the New York Department of Insurance introduced Regulation 126 several years ago. This regulation has served as the "groundwork for a lot of the early testing and still serves as a focus point of the testing done today," Barry says. "For the most part, all life companies must submit an actuarial memorandum as part of their annual statement with the regulators as it relates to Reg 126."

The core of the actuarial memorandum is a deterministic testing of seven different interest rate scenarios. These are:

We do try to understand the industry standards, the third party software that is out there. We do visit vendors to understand the models and the main drivers of the models. One of our conclusions is that the models are very sensitive to certain assumptions, many of which are provided by the companies themselves.

Another thing to focus on is the internal hedges that may be in place in a diverse product portfolio. For instance, a company that writes single premium deferred annuities (SPDAs) as well as a lot of single premium immediate annuities has a natural interest rate hedge.

Also, we try to understand the insurers' risk appetite and whether or not their analysis of risk is consistent with ours. We certainly do like to see senior people in the organization that have a good understanding of the various risks, as opposed to the risk management being solely an actuarial department function. Most of our work is done in the context of our review meetings with management and through follow-up meetings, when necessary, with the appropriate personnel.

I would say too that we recognize analysis of interest rate risk and ALM management is probably the most difficult part of examining a life insurance company. So we're always trying to upgrade our knowledge and capabilities. But just to reemphasize, the key to understanding is really on the product side - the product mix and how and to whom the products were sold.