Corporate Power Tool.

Asset/liability management can help executives make better decisions about what products to bring to market, how to cover their risks and what markets to enter.

Chief executive officers want to make the best decisions for their businesses--decisions about what products to bring to market, how to cover their risks and bets, what markets to enter or exit. To make good decisions, they need an edge--a tool that will make them faster, smarter and more powerful than their competitors.

Asset/liability management, or AIM, has looked to many CEOs exactly like that kind of tool--and for good reason. Asset/liability management, when used appropriately, has helped many companies and their CEOs make better strategic decisions. But other executives who have tried to use asset/liability management have not enjoyed the success they had anticipated. Instead, they have gone through a kind of day-after-Christmas funk: Once the newness of asset/liability management had worn off, they were disappointed. They didn't get the results they were looking for, or the results they received came at too high a price. Some gained a great deal of information, but it didn't seem to mean anything useful. Because of this disappointing experience, these CEOs put asset/liability management aside and began to look for the next tool.

That's unfortunate and wasteful, because asset/Liability management is a very good tool. But as with any other tool, it is most valuable when the user knows what it is best used for; what, if any, parts to outsource; and how to evaluate proposals to use it.

Best Uses for ALM

To determine the best uses for asset/liability management, it is important first to understand what it is. Asset/liability management is the umbrella term for a set of processes and tools that senior management can use to make decisions aimed at improving shareholder value. The processes, which often go under the banner of integrated risk and capital management, support management decision-making regarding capital allocation, product pricing, reinsurance, asset allocation and capital structure. These processes consider the effect of any one strategy in light of the other strategies pursued. They also consider the effects of the combined strategies on the key financial statements of the enterprise. Software tools support the required modeling of the enterprise. The tools often are known in the insurance industry as "dynamic financial analysis"; outside insurance, they're known as large-scale simulation. While the term "asset/liability management" at times is applied to a software tool alone, it is more accura te to think of it as the combined set of processes and tools.

Asset/liability management sometimes has been called "the actuary's spreadsheet," because it is essentially a complex and comprehensive financial model of the institution. The model simulates the balance sheet, income statement and cash-flow statement. The model takes into account such things as future revenues, future expenses, claims, investment returns and capital structure of the enterprise. Furthermore, it does so in a large-scale simulation so that you can see the range of possible outcomes and the likelihood that they will occur.

An additional source of power for the model is that it can calculate effects on both sides of the balance sheet from such events as future sales, changes in interest rates, inflation, currency valuations and claims occurrences. That is, it is capable of taking a very long view of the enterprise and its Environment--the view of the CEO. That makes it distinct from other, short-term analyses used to manage assets and liabilities, such as value-at-risk analysis.

As a comprehensive and sophisticated model, asset/liability management is best used to answer detailed strategic questions, such as:

* What can I do to improve my return on equity?

* Do I have adequate capital to manage enterprisewide risks--both product risks and investment risk?

* Is this merger likely to create value?

* How can I protect my balance sheet against catastrophes?

* What product mix should I offer?

Because it is a sophisticated modeling tool, asset/liability management should not be used to answer questions where the cost of the analysis exceeds the benefit. For example, it should not be used if:

* the decision is operational;

* the likely returns are small compared with the effort required to calculate various exposures; or

* the user is constrained from making changes in strategy because of other considerations.

CEOs would not want to go to the trouble and expense of using asset/liability management modeling if company assets exceeded liabilities so much that the company didn't need to slice its investment strategies too finely. For example, a company with $32 million in assets and $4 million in incurred-but-not-reported claims can be quite aggressive on its asset mix.

An executive would not want to run a full asset/liability management analysis to decide whether to take on a small but standard risk. For example, a large property/casualty company with many policies covering truck fleets doesn't need to perform this kind of analysis to see if it makes sense to cover a new customer with a fleet of five trucks. (On the other hand, a company could use asset/liability management to decide the larger question of whether it wants to be in that business at all.)

CEOs also would not need to perform asset/liability management analyses if they knew that they were constrained by other considerations, such as their basic value proposition. For example, if a financial-services company's proposition to customers is "safety and security" because it invests in bonds, then it couldn't invest in equities, because it would be telling the market one thing and actually doing another. A company does not need a sophisticated asset/liability management analysis of asset allocations to make that decision. (That said, a financial-services firm could use such an analysis to determine the opportunity cost of basing its strategy in the first place on offering a conservative balance sheet to the market.)

Build It or Buy It

Because asset/liability management modeling is very sophisticated, the "sticker price" from an outside vendor can look high. A company must consider whether to build its own model or pay this price. The do-it-yourself cost can look lower because the people resources already exist.

In making the decision, a company should consider these questions: Do internal resources have the adequate capabilities and is the opportunity cost of dedicating resources to building the model commensurate with benefit?

Answering the capabilities question may be the most important. The capabilities required to create this "actuary's spreadsheet" include the ability to model over long periods such things as:

* product sales behavior;

* product risk;

* investment asset risk, from all classes of assets, including equities, domestic bonds, foreign bonds, corporate bonds and mortgages;

* currency risk;

* inflation; and

* tax and accounting treatment.

Most companies find that they don't have full range of capabilities within their organization. If they do, they likely reside in a number of different parts of the company. They also find that even if they have the capabilities, they don't have equal competency in all of them. For instance, most property/casualty insurance organizations are generally much better at modeling liabilities than assets. The result of uneven capabilities is an "unbalanced" total model. The liabilities part of the model may be quite robust; but the asset part of the model may be too simple. That can lead to fallacious findings.

Finally, even if the necessary capabilities and competencies exist on staff, most companies find that the opportunity cost of dedicating staff to building the model is prohibitive. It is roughly the equivalent of deciding to answer a simple net present value question by deciding first to build a spreadsheet program.

All that said, building a model can make sense if both sides of the model are balanced and the model is fairly simple. To assure the balance, a company could build the parts for which the organization is most competent--such as liability modeling--and buy the other parts "off the shelf." For instance, if a company is using asset/liability management for product pricing and it is going to invest in mortgages, it can outsource the modeling of a portfolio of collateralized mortgage obligations using a product such as BondEdge, a widely used analytical tool for fixed-income investments. A company can develop a great deal of intuition about its business in the act of building fairly simple models, which can be a precursor to something more sophisticated.

When to Use It

Asset/liability management is like any other new tool in that people will be tempted to use it without thinking through whether it makes sense to do so. That's a bit like a do-it-yourselfer deciding to use a new hydraulic, 2,400-horsepower nail driver every time he needs to drive a nail--whether putting up studs in a new recreation room or hanging a picture.

Using asset/liability management to evaluate proposals--whether internally or externally generated--companies should consider these questions:

* Does the proposal respond to a regulatory mandate, such as for cash-flow testing? If so, a company has no choice and must go forward. The temptation in this case will be to do the minimum amount of required work at the lowest possible cost. Companies that do this may be missing an opportunity to use the requirement to answer more fundamental and strategic questions about their business.

* Does the proposal have a clear objective? This is the most important question. The objective should be precise, specific and tangible with a set deadline. For example, "Our competitors tend to have a higher equity allocation than we have and have added a lot of value to their organization. Does it make sense for us to increase our equity exposure to match theirs?"

* Is the proposal logically consistent? That is, does the degree of complexity and sophistication of the means--building an asset/liability management model--match the degree of complexity and sophistication of the end? For example, wanting to know the impact of buying a high-retention insurance contract does not require using an asset/liability model--sometimes the actuary's spreadsheet need be only a spreadsheet.

* Is the proposal "balanced"? That is, will the model be built by a team with adequate breadth and depth of capabilities and competencies? If it is strong on one side and weak on another, then it will only be as strong as its weakest link.

* Is the proposal promoting the purchase of a "black box"? That is, can the model builders give reasonable answers to reasonable questions, such as: What will it model? How will the model work? What are the assumptions? If the model builders cannot give good answers to these questions, then there is high risk the model won't be reliable.

Asset/liability management is an effective tool to guide strategy. It has been used in asset-allocation reinsurance to work out buying blocks of business. It offers a full financial model of the organization and it integrates asset and liability risks. To get full value from it, companies must keep in mind the kinds of decisions most appropriate for it; when it makes sense to "do it yourself" and when to go to experienced outside experts; and how to evaluate proposals for using it.

Stephen Britt is a consultant with Tillingbast-Towers Perrin, Hartford, Conn.