A custom mix: a growing number of insurers are adding product liability considerations to create a stronger investment management process.

Key Points

* Asset-liability modeling is the most crucial element in the integrated investment-management process.

* Life insurers and P/C insurers tend to focus their efforts on different components of the process.

* If assets may not be profitably invested for certain products within enterprise risk constraints, pricing assumptions are changed despite competitive pressures.

Increases in insurance industry competition have resulted in tighter premium margins and greater product optionalities. As a result, the importance of asset management within the industry also has skyrocketed: greater product optionality makes liability modeling far more complex than a decade ago, and tighter premium margins result in smaller asset reserves to cover instances when actuarial assumptions were too aggressive. Under pressure for profitability and facing the low-interest-rate environment of the past decade, many insurers have migrated to riskier assets in search of higher yields.

In the current environment, integrating liabilities into asset management is critical to risk optimization. Leaders within the industry have begun to recognize the greater risks in their asset-management practices, and select firms have been active in integrating product-liability considerations into their investment strategies, and reasonable, risk-adjusted return projections into product design.

To illuminate the current state of insurance investment management and to bring a disciplined approach to an understudied field, Patpatia & Associates undertook a comprehensive survey of insurers to identify the market trends and evolution toward integrating asset-liability management within general-account asset-management strategies.

We spoke with 53 firms varying in discipline type (life, including annuities and health, and property/casualty) and size.

The results suggest that across the industry, insurers have not placed significant emphasis on investment function. It is not that insurers' investment departments lack depth of knowledge about investments. Insurance companies are some of the largest institutional investors, frequently with tens or hundreds of professionals. Rather, insurance investment departments typically have been run separately from the products being sold, with a primary focus on maximizing investment income--the basis on which corporate management typically measures asset-manager performance. Management has treated the investment department as a "black box," counted upon to produce its allotted share of the bottom line--but otherwise allowed to self-manage in its own silo. This contrasts dramatically with other financial services businesses, such as banking and pension funds, where proactive asset management is the principal strategic function of the industry.

In design, asset-liability management departments are supposed to bridge the disjointed relationship between product management and investments. However, in most organizations, the ALM department seems to focus on regulatory reporting rather than managing the liability and asset exposures for product management and investment management.

Liability-Driven Investing

Sixteen of the 53 firms Patpatia & Associates surveyed were at the forefront of industry investment management. All had developed investment-management techniques that incorporated liability metrics into their investment policy and benchmark process. Six broad components were identified as core to successful investment policy: liability modeling, asset maturity profiling, liquidity modeling, asset allocation, benchmark creation and returns analysis. Implementation of these processes is central to liability-driven benchmarking.

Effective, well-integrated asset-liability modeling, driving a comprehensive portfolio strategy clearly underpins the entire benchmarking process. Individual insurance contracts are modeled for their anticipated cash flow based on assumptions of their characteristics--product features and optionalities. These liability metrics then are matched with a portfolio of assets to determine the optimal investment policy to best match the required payments while providing a maximum economic return. A sensitivity analysis across multiple economic scenarios ensures that within a reasonable market environment, assets will provide the necessary cash flows to meet liability payments.

Many insurers continue to develop investment targets (frequently duration alone) based on the intuitive experience of the investment department. However, more than half of all survey respondents indicated that duration alone is an insufficient ALM metric and that cash-flow profiles done at least quarterly also are needed to develop an effective investment strategy. Liability cash-flow profiles show the nuances in anticipated payments that duration and convexity cannot reflect, and they serve as the foundation for an asset maturity structure. This ensures a cash-flow match between assets and liabilities, reducing required portfolio liquidations, capital losses and volatility of earnings.

Most insurers employ two discrete systems for asset-liability modeling: an ALM system to model the liabilities and run the combined cash-flow match, and a fixed-income analytic system to generate asset cash flow.

Of the insurers implementing or about to deploy liability-driven benchmarking, most had 100 to 150 discrete liability portfolios, and the majority of respondents model liability at the product-group and business-line level, rather than for each individual portfolio. The average number of unique liability models was 20 to 25, with the level of liability modeling determined primarily by the size, liability characteristics and legal jurisdictions of insurers.

Liability cash flows then are translated into an optimized maturity term structure maximizing free cash flow (asset maturity profiling). Assumptions then are stress tested to model potential downside risks and set liquidity requirements (liquidity modeling). Based on these inputs, strategic and tactical asset allocations may be developed (asset allocation).The investment strategy then may be formalized in explicit benchmarks (benchmark creation) against which performance can be measured and results attributed (returns analysis).

Differences in liability-driven benchmarking are apparent across insurance lines. Life insurers focused on liability modeling, asset maturity profiling and asset allocation when developing investment policies and benchmarks due to predictability of liability portfolios. Property/casualty insurers tended toward liquidity modeling driving the asset-allocation process, as their respective liability portfolios generally lacked the predictability required to set complicated maturity term structures.

A Common Language

Within a system of integrated investment management, the investment benchmark is the tangible incarnation of the asset-liability strategy. It provides a common language for product management and the investment department to communicate their needs in the process. The benchmark represents the interdependent relationship between liabilities and assets, as liability payment streams dictate the required asset cashflow maturity structure.

At the heart of this approach are personnel and risk-modeling technologies. Improved risk budgeting also is required, fostering a diversification of exposures, rather than focusing on one or two risks where the investment department feels it can outperform the market by private placements, a frequent insurer practice. Critically, if assets may not be invested profitably for certain products within enterprise risk constraints, pricing assumptions are changed despite competitive pressures.

The benefits to this approach go beyond liability risk mitigation. Survey respondents employing liability-driven benchmarking reported additional benefits, including improved comparability and oversight of portfolio managers, yield enhancement and loss mitigation, and the ability to incent portfolio managers based on liability demands.

Obstacles to Implementation

Although a majority of the 53 respondents believed liability-driven benchmarking was the optimal means to manage insurance assets, execution complexity thus far has prevented a majority of large insurers from implementing such a system for asset management, while many midsize to small insurers indicated an overall lack of familiarity with both the process and inherent benefits of deploying liability-based benchmarks. Significantly, many mutual insurers often lacked a liability-based focus. With a few notable exceptions, their reduced need to emphasize stable quarterly earnings has resulted in more basic benchmarking processes--duration only, for example--and weaker ALM risk controls.

This is unlikely to remain the state of the industry, however. Forty-three percent of respondents indicated they were reviewing, implementing or recently had completed enhancements to their investment-benchmarking processes to integrate liability metrics into their investment policy and benchmark processes. This shift is a result of insurers recognizing the increasingly competitive landscape that has forced premium prices down and increased the complexity of products offered. These new risks, in combination with riskier assets (lower quality credits, duration, illiquid securities, etc.) suggest that without liability-driven benchmarking, corporate profits likely will be affected.

Integrated Investment Management

Successful investment policy requires the following components:

* Liability Modeling

* Asset Maturity Profiling

* Liquidity Modeling

* Asset Allocation

* Benchmark Creation

* Return Analysis

Contributor Mark Chin is senior business analyst with strategic consulting firm Patpatia &Associates. He may be reached at [e-mail].

Important Components

Life insurers and property/casualty insurers that responded to the
Patpatia & Associates survey perceived the following components
as highly important to the liability-based benchmarking process

                   Life *        Property/Casualty

Liability
Modeling            82%                56%

Asset
Maturity
Profiling           64%                22%

Liquidity
Modeling            45%                78%

Asset
Allocation          73%                89%

Benchmark
Creation           100%               100%

Returns
Analysis            91%                78%

* Including annuities and health

Note: Table made from bar graph.

Reaping the Rewards

Respondents to the survey named the following as
benefits of liability-based benchmarking:

Improved comparability and oversight of portfolio
managers/third-party managers                          79%

Yield enhancement and loss mitigation                  71%

Portfolio managers incented to meet demands of
individual liabilities                                 71%

Improved stability of earnings with reduced
need for unplanned liquidations                        64%

Quantification of risk limits and
out-of-tolerance exposures                             57%

Business line/product-level performance
and profit measures                                    50%

Investment policy independent of portfolio
manager's preferences/specialties                      36%

Reflection of pricing constraints within benchmarks    29%

Investment Survey Participants

Insurance companies that participated in Patpatia & Associates'
investment-management survey represented various sizes, types and
business lines.

Size of Company

Assets Under
Management
<$1 billion                   45%

Assets Under
Management
<$50 billion                  25%

Assets Under
Management
<$1 billion-$50 billion       32%

Type of Company

Mutual                        36%

Public                        64%

Business Lines

Life and P/C                   7%
Property/Casualty             36%
Life                          57%

Note: Table made from pie chart.