Stretch Management for Insurance Companies

Stretch Management for Insurance Companies

The author of this article discusses whether stretch management is a strategic tool or a tactical trap.

By James Ramenda

Management using stretch goal-setting is a concept rooted in the corporate lore of one of the world's largest companies, General Electric, and championed by its legendary former CEO, Jack Welch. There's no formal definition for stretch, but according to Welch it "... essentially means using dreams to set business targets with no idea how to get there."However ethereal this may seem, it is the starting point for decisions that have tangible consequences for a company. For example, a company with an ROE of 10 percent may look at the profitability of a broader corporate universe earning, say 15 percent, and decide that it must improve to this level by the end of five years to justify its use of shareholders' capital. A change of this magnitude is easy to say, hard to do–precisely the point of stretch.

Certainly, no insurance company could effect such a change without drastically changing the way it does business. The stretch goal forces the company to evaluate its business platform and the levers available to it in a completely new light. Levers that are most readily within reach (e.g., new pricing/underwriting practices, new versions of existing products) typically won't achieve a stretch goal. At a minimum, some levers that are beyond immediate reach (new product lines, new distribution and new markets) must be considered and even then, may well fall short. More likely, a change of stretch magnitude will require not just new levers, but a new platform (new capital structure, entry into new lines of business, exit from others, acquisitions and/or divestitures). And, of course, it may also be that some of the levers/platforms needed are simply too great a stretch, at which point the company may decide to put itself up for sale. Any of these decisions obviously will have major tangible effects on the company, its structure and its people.

Like companies in other industries, many insurers indeed adopted the principles of stretch (whether formally using this term or not) over the past two decades and effectively reinvented themselves. Experience in advising and managing investments in insurance companies has led me to conclude stretch can either be a very useful tool, or an unfortunate trap for insurance companies, depending upon how and for what purpose it's applied. Specifically, stretch is best-suited to long–term strategic planning and particularly ill–suited to short–term, year–to–year goal–setting. As a case study, let's consider the industry experience with variable products.


If we go back to the 1990s, that decade began with life insurers' profitability facing a number of pressures.

Large amounts of interest–sensitive and universal life products written in the previous decade had seasoned, with some disappointing realizations. Generally, profit margins had not lived up to their billing, as interest rate spreads were subject to intense competitive pressure. These pressures also led some companies to court investment risks that proved difficult to manage, creating income statement and balance sheet volatility.

Accounting standards were introduced (FAS 97 and FAS 115) that increased the focus on these narrow margins and increased volatility.

Regulators were becoming increasingly sophisticated in the determination of risk–based capital requirements, while rating agencies steadily raised the bar for top ratings.

The pool of replaceable traditional policies had dried up and new markets needed to be opened.

At this time many insurers were earning ROEs in the range of roughly 9– to 12–percent, too low in the view of some shareholders and boards, especially as they were incurring more volatility than these returns seemed to justify. As a result, some fundamental stretch questions were asked:

  • Can we get to a meaningfully higher ROE with our current business model?
  • Can we do so without increasing the risks and capital requirements of the business?
  • If not, what businesses should we be in?

In response, a number of insurers successfully applied the concept of stretch over the 1990s. A major realignment took place with many companies radically changing their business mix. For example, nearly all but the largest players exited the group medical insurance line. Several multi–line companies became either life/health or property/casualty companies, but no longer both. Consolidation accelerated and a trend toward demutualization began to take hold. It's noteworthy that among the major restructurings was the insurance arm of Welch's General Electric.

Another notable change was the decision by many individual life and annuity companies to pursue variable products as a core line of business. Variable products were conceived and introduced quite a few years before they became a major product for many insurers, but apart from some well–defined niches, their general popularity tended to ebb and flow with the stock market. By the mid–1990s, the forces of the market, the economy, and the accounting/regulatory environment combined to bring the strategic advantages of these products into focus. Rather than try to maintain interest–sensitive product spreads and tolerate investment risk, insurers could instead earn fee–oriented income while the policyholder would bear the bulk of the investment risk. The initial result was steadier income (and steadily increasing income during the bull market years) and less required capital, the best of both worlds–exactly the kind of dream results that Welch's stretch is intended to produce.


Fed by the bull market, variable product lines produced double–digit asset and earnings growth for several companies in the 1990s. But as all businesses do, the variable business began to mature as more companies entered and the target market became more penetrated. Two additional events added pressure to the business:

  1. The slowing economy and bear market of 2000-2002 brought investors back to the realization that markets can go down (and sharply), as well as up.
  2. The reduction in long–term capital gains tax rates, a policy response to the economic slowdown, mitigated a key advantage of annuity products, i.e., the tax–deferred inside build–up of income.

This occurred against a backdrop of an industry that had successfully stretched in many ways during the previous decade. Indeed, by that time, stretch found its way into the lexicon of the annual planning cycle. For example, a company might target a growth rate of 10 percent, with a stretch goal of 15 percent. Or some managements might just skip the distinction and target numbers well above trend line with the logic that while they probably wouldn't get there, they'd probably get a better result simply by setting a goal with some stretch in it.

For some variable companies, the unintended, but by no means immediate, result of this short–term stretch thinking was to reverse the long–term strategy underlying the move to variable products. Dealing in the short term, they pulled the levers within their reach, i.e., pricing and product design. Companies needed product differentiation to continue their high rates of growth. They also needed to give annuity holders comfort that their nest eggs wouldn't be wiped out by the next market downturn. Variable products became increasingly "de–variable–ized" by the introduction of more and more aggressive guarantees and ratchets. These features guaranteed and/or built withdrawal benefits, death benefits, and income benefits decades into the future.

The strategic reversal was virtually 180 degrees. A product that was initially adopted in part to insulate the company from investment risks now became a long–term investment guarantee product. When the financial crisis hit (itself a product of short–term stretch management in the banking sector), the decline in asset values coupled with the impact of guarantees embedded in many variable products led to a reversal of the strategic stretch goals: Profits turned to losses, volatility increased beyond any expectations, and some companies were forced to consider government bailouts.

The lever within reach in an annual planning cycle, in this case product redesign, was not adequate to produce extraordinary growth while maintaining profitability and risk characteristics. By definition stretch requires enough time to find new levers, even new platforms. Anything immediately within reach won't do it.


Growth targets like those discussed above, i.e., 10– to 15–percent, hardly would be shocking to anyone involved in strategic or annual planning. Yet, these rates of growth are multiples of the 5– to 6–percent long–term growth rates in U.S. corporate earnings (same for nominal GDP). One must question how it is that companies can routinely and consistently hope to achieve growth that is two or more times the growth rate of the overall economy.

Of course, parts of the answer are not completely unique to variable products or even insurance. There's genuine confidence, also maybe hope, as well as the reflexive response to pressure from shareholders, analysts and boards of directors who desire aggressive management. But for insurers, there are some specific factors relating to growth and volume that tend to lead them directly into the short–term stretch trap.

The insurance industry has fought long and hard to improve its expense efficiency. Cuts are painful. The prospect of growing into one's existing expense structure is a more attractive and dynamic solution.

Insurance markets are often nearly perfectly competitive, so distributors have plenty of choices among product manufacturers and consequently plenty of power in insurers' decision–making. This tends to give sales volume a greater weight in product pricing and design decisions.

The levers for building volume that are within reach over a short–term planning cycle basically are price, commissions, underwriting standards and coverage features. Pulling these levers is painless in the short–term because the net effect will not be measurable for many years, perhaps taking a complete business cycle, or some other economic event, to fully develop.

It is this last point that most clearly distinguishes insurance from many other businesses, particularly those in the real sector. An insurer's short–term mistakes can last for decades. Conversely, if a consumer product company like, say Nike, misses its volume goal, it's an unsold inventory problem, but the bottom line impact is short term. More sneakers hit the discount racks. But the customers that buy their sneakers on sale don't get to keep buying them at that sale price for the next 30 years. And there's no contract saying they can recover the purchase price many, many times over upon certain events. It really doesn't matter much to Nike who buys its sneakers and more is better. For insurers, it matters a great deal who buys their products and more can be much worse, and for a long time.

Of course, disciplined companies can navigate these pressures, even if it means the occasional short–term slowdown in growth. This can be disappointing to Wall Street or the board of directors. The word unexciting is sometimes used. But insurers are in the business of giving long odds. A sudden rush of excitement, when giving odds, is usually not a good thing. What is exciting about insurance happens over the long term: Earnings compound because sales compound premiums, premiums compound assets, and assets compound with interest.


Stretch poses several challenges for the actuary/CERA acting in a pricing or risk management capacity. The immediate challenge is technical. Stretch is intended to take the company beyond its apparent limits while having, as Welch writes, "no idea how to get there." This may mean making product or underwriting changes without credible experience upon which to draw. The limitations or outright absence of such data cuts two ways. Obviously, the actuary must rely on judgment to a greater degree (maybe entirely) than where data is plentiful. More subtly, those encouraging the stretch can point to the absence of such data as diminishing the relative importance of the actuarial conclusions, or even being an advantage, i.e., the first mover advantage in a product area. Akin to trying to prove a negative, it is a conundrum for an actuary to say "no" to the field, or "no" to one's boss, while admitting to having little or no data.

A second dimension to the challenge is that even if emerging experience matches management's expectations, this is in no way indicative of tail risk. Whether tails are fat or not, it may be years before a significant tail event occurs, if indeed one ever does. The accumulation of data without a tail event increases the sense of trying to prove a negative. As years pass, those who warn about the risk may seem more and more divorced from reality.

Compounding these challenges is the difference between long–term and short–term stretch. Ironically, short–term stretch may seem less risky because it deals with incremental movements of reachable levers such as price, underwriting standards, etc., while long–term stretch involves changing the business platform through big–ticket items, e.g., acquisitions, divestitures or other restructurings. Boards of directors will rightly ponder the latter in great detail, while typically leaving product changes to the line of business managers to decide.

The greatest challenge, and crowning irony, may occur if the short–term stretch goal is surpassed on the strength of a surge in business sold. Because insurance markets are generally highly competitive and fragmented, an incremental change in product can produce a flood of business. While this is welcome news for companies in most industries (back to the Nike example earlier in this article), it can be a dire warning for an insurer. Examples abound in insurance of extraordinary growth being a warning sign: early versions of long–term care, workers' compensation carve–out business, and as discussed here, the rapid spread of generous guarantee and ratchet features in variable products. However, the possibility that a rapidly growing line of business could be a threat is highly counterintuitive to business people from other industries (i.e., boards of directors) and even some insurance managers. In this case, warning about tail risk may seem more than divorced from reality; it may seem to be running in the opposite direction.

So for the actuary, stretch poses a multi–year challenge no matter what the planning horizon, including:

  • Identifying the risks in the stretch goal.
  • Quantifying the risks, even in the absence of credible data, in ways that non–technical management can appreciate, i.e., identify when one is betting the company.
  • Maintaining awareness of the degree of risk even as it fails to materialize and the business grows with seemingly ever–greater success.

The process of risk management in a stretch environment dovetails with behavioral economics. Executive management and boards of directors are comprised of highly successful individuals. To get where they are, they've had to outperform stiff competition, often by finding opportunities that others missed and/or by taking risks that paid off. They admire highly successful (and well–publicized) management strategies in all industries and may seek to emulate them, regardless of the differences among industries. They typically place great value on personal confidence, an enthusiastic presentation, and a can–do–the–impossible attitude. In short, they tend to believe deeply in stretch in all its forms. Presenting a quantitative analysis of risks that may be only 5 percent (or less) likely to occur over a span of decades doesn't play well to this predisposition.


To avoid the short–term trap that stretch may pose, the actuary/CERA must recognize that more than analytical tools are needed to bridge the gaps that may exist between these approaches and the real–world experiences of many stretch advocates. The good news is that believers in stretch generally are also firm believers in markets. Obviously, the financial markets of the past two years have given a huge, albeit costly, boost to risk awareness. This crisis provides case studies in nearly every industry. Indeed, the center of the meltdown, the aggressive efforts to reach new customers (sub–prime) by lenders, is a perfect example of how the short–term stretch trap works. The effort to reach a new market (sub–prime) failed to take full account of the differences in this new market, differences that were obscured by rapid penetration in a period of economic prosperity.

Certainly, many case studies from history in insurance stretch performance can be instructive, among them:

  • The rise and fall of SPDA companies in the late 1970s and early 1980s.
  • The bubble–like collapse of the commercial mortgage–supported GIC market a decade later.
  • The underpricing of early forms of long–term care insurance, a much heralded growth opportunity for the industry at the time.
  • Market conduct issues relating to illustrations of highly popular universal and interest–sensitive life products and features like "vanishing premiums" which sometimes failed.
  • The sudden bubble and nearly simultaneous burst in workers' compensation carve–out business in the 1990s.

All of these should be required reading for insurance company division heads, CEOs and directors, and there are many more examples.

The variable product case study presented here may be particularly instructive because it shows both sides of the stretch process, the long–term strategic goals that were envisioned and initially realized, and the short–term tactical thinking that subsequently reversed the outcome for many companies.

Recognizing that stretch can be either good or bad is an important point for the actuary/CERA, particularly in light of the behavioral economics discussed earlier. Relying on analytics and/or negative case studies to highlight risk is incomplete and may even be counterproductive if perceived to be part of an inevitable nay–saying exercise. But if these tools can be augmented with an effort to salvage a potentially flawed initiative, the process will be viewed more positively and may also provide needed clarity along the way. For example, if a given tail risk has highly conjectural severity, a market price may be developed by soliciting reinsurance or capital markets solutions. If the tail risk can be shared or transferred at reasonable terms, the overall economic profile changes dramatically. Alternatively, if there is no market for the risk, then this fact is very telling for everyone around the table. The very process of attempting to mitigate the tail risk will focus attention where it belongs–on the nature of the risk and not the constituencies advocating for or against.

Stretch goal–setting is integral to the planning process as practiced by many business people, whether formally recognized as such or simply implicit in their thinking. The nature of insurance, however, requires that great care be used in applying stretch. For the actuary/CERA, this means not avoiding stretch, but as the variable product experience shows, recognizing when it is strategically necessary and when it is tactically unwise.

James Ramenda, FSA, CERA, is managing director for Northington Partners, Inc. He can be contacted at