Select Language

Check Application Status

Resource Zone

What life insurance illustrations don’t tell you

Richard M. Weber, MBA, CLU

Rate 1 Rate 2 Rate 3 Rate 4 Rate 5 0 Ratings Choose a rating
Please Login or Become A Member for additional features

Other formats

Audio 0:59:09

Note: Any content shared is only viewable to MDRT members.

Since the introduction of paradigm-shifting universal life policies in the late 1970s, producers and clients have had to reconcile optimistic sales illustrations with the in-force reality of values that are often less than expected. Forty years later, we still struggle with illustration disclaimers. Weber demonstrates new ways to advise clients about making current assumption life insurance policies meet their expectations and be in their best interest. This process diagnoses new and in-force IUL and VUL policies and explains how to discuss best (not lowest) premiums and outcomes. You will learn how to achieve consistency with the client’s expectations and conduct a conversation with a client who already owns a current assumption policy.

Click here to find more from the 2018 Annual Meeting

When I’m introduced as being a 50-year veteran of the life insurance industry, I have to quickly add “I was a child agent!” to avoid being immediately thrust into the category of “old fogy.” Well, almost. If that many years means anything, hopefully it includes a smidgen of wisdom, a soupçon of expertise, as well as a whole lot of “well, in my day . . .” and maybe just a pinch of “I told you so!”

Consider the ultimate oxymoron: I was a junior at Cal Berkeley in the mid-1960s; I was in Navy ROTC; and I was accepted into Provident Mutual’s campus internship program. There I was paid a modest $50 a month to go through some rudimentary training to see if selling life insurance was something I might be good at. Trust me: Selling life insurance was nowhere near the top of my list of careers. And yet there was something about it that resonated with me—being in business for myself, no limit to the amount of money I could make, and having the opportunity to be in a helping relationship with the many people who first were clients and then became friends.

I’ll end this brief reminiscence with the recollection that became especially significant for me in the 1980s with the surge in excitement about universal life policies (or at least their illustrations), while, at the same time, much of the industry was divorcing itself from their career agents and demutualizing. The reminiscence is that the first thing I learned in that campus training program 50 years ago was that “mutual is the way God intended life insurance!”

I’ve already named the two most disruptive events in the industry in my 50 years: the introduction of current assumption policies (universal life) and abandonment by many major companies of the single-focused business model of “mutual” as a means of enhancing policyholder value. And, yet, thinking about it now, I see that it was inevitable: With the disruptive force accompanying the introduction of policies whose main feature was “pay what you want, when you want,” the then-220-year-old economic model of traditional life insurance was broken. Supporting career agents is an expensive and long-term proposition, long supported by the mainstay product of whole life insurance, but which could not be supported by the uncertainty of unpredictable premium flow and the use of computerized policy illustrations to demonstrate that “my premium” is lower than “their premium.” And, of course, underlying that drive to the lowest and best price was the ironic observation: “If it’s spit out by a laser printer, it must be true!”

This year marks the 40th anniversary of the introduction of the first generation of universal life, anecdotally accomplished by a small carrier just a few miles down the road from here called E. F. Hutton Life. UL was a paradigm shift in an industry that hates paradigm shifts! UL was transparent! UL was unbundled! UL was paying crediting rates—for a brief moment in time at its introduction—of as much as 14 percent. And, in a brilliant stroke of timing, it was introduced at the dawn of the desktop personal computer, without which UL would have been impossible to sell. The success of its attraction to consumers and agents was nothing short of startling. Where whole life had an 88 percent market share by premium in 1978, that dominance dropped by more than half by the mid-1980s, with a percentage point by percentage point capture of that falling dominance in market share taken over by universal life. Even as interest rates began their historic decline in the economy, from nearly 16 percent for the 10-year Treasury bond in December 1980, the profound shift from traditional to universal life was “on” even as UL crediting rates begin creeping down to 10 percent by 1985.

It turns out that the interest rate decline wasn’t temporary; exacerbated by the economic events leading up to the Great Recession in 2008, the slide of 10-year US Treasury bond rates from that mountaintop of 16 percent had plunged to only slightly more than 1 percent by the end of the millennium’s first decade.

Looking back, then, it’s easier to see UL as a life insurance concept and product with a very attractive appeal to agents, with its evolving four generations of crediting rate design as well as appeal to their clients. As declining interest rates in the economy continued to push generation one UL crediting rates noticeably downward by the late 1980s, and with the rise of a bull stock market that would ultimately return a compound average of more than 18 percent a year by the time it relented in early 2000, variable UL—generation two—became an equally hot successor to the original design in which crediting rates were set by the carrier’s board of directors.

As UL products continued to reflect the changing economy, we progressed to so-called no-lapse, or guaranteed death benefit, UL, for those who were concerned the stock market couldn’t go up forever—what I call the third-generation product. And, finally, we get to generation four—index UL—largely propelled into the limelight by “zero is the hero” marketing slogans following the extraordinary drop in stock market values between November 2007 and March 2009, along with interest rates that were barely above 0 percent on even one-year commitments such as CDs.

Different in reasons but similar in results of the degree to which the original, high, current interest rate environment of the 1980s contributed to a substantial amount of disintermediation between whole life and traditional UL, index UL began building market steam as the stock market began to recover in spring 2009. While older blocks of traditional UL were only earning the contractual guaranteed crediting rates, IUL was propelled by its unique design of deploying premiums in excess of a 0 percent guarantee to hedge the stock market and provide at least partial participation in the upside of a one-year index, such as the S&P 500. It was an extraordinary variation on variable UL, without the downside and without the need to have a securities license to sell it.

In its first successful iterations, IUL illustrations portrayed the assumption of constant crediting rates in excess of 8 percent, along with the storied guarantee of no worse than 0 percent. It was a perfect launch into the bull market that resumed in early 2009.

But, as tends to be the case, sales illustrations with very high assumed crediting rates prompted sales enthusiasm about this new UL design to de-emphasize the value of death protection as the primary purpose of life insurance and move more to take advantage of the unique income tax advantages of life insurance—the tax deferral on the inside buildup of cash value, tax-free withdrawals to basis, tax-free loans beyond basis—all of which is forgiven when the policy becomes a death benefit.

By 2014, the Wall Street Journal took notice of the New York Insurance Department’s review of IUL illustrations and a concern about overstated return possibilities in sales illustrations.

This should have (but didn’t) caused the industry to self-regulate and begin to find ways to suggest advantages of this new UL design without seemingly “promising the moon.”

Because the real question should have focused, from the very beginning of universal life sales illustrations, on “what crediting rate should be used to calculate planned premiums” in response to the understandable question from a client: “What’s it gonna cost?”

And especially when we know (or should have known), as pointed out in 1992 by the Society of Actuaries in the face of the first universal life “illustration rate” crisis leading to illustration regulation, the following:

  • “There is no simple measure or analysis to compare future performance of unpredictable events. This fact is well understood in the securities industry and needs to be assimilated into the life insurance industry as well.”
  • “How credible are any non-guaranteed numbers projected 20 years in the future, even if constructed with integrity?”
  • “Most illustration problems arise because the illustrations create the illusion that the insurance company knows what will happen in the future and that this knowledge has been used to create the illustration.”

In 2015, regulators stepped in and adopted an Actuarial Guideline (AG 49) to limit the crediting rates that could be used on IUL sales illustrations to essentially reflect a 25-year rolling average return under the illustrated index, most popularly the one-year point-to-point S&P 500 (without dividends).

But consider: An essential problem with policy illustrations from the very beginning of UL projections was the fact that whatever “number” was allowed by the insurance company’s software to serve as an illustrated crediting rate, that number (and typically higher rather than lower) would be used as a constant for the 50 to 90 years of projected values out to the maturity of the policy—up to age 125 or even longer.

For VUL and IUL policies whose crediting rate is ultimately determined by the performance of “the market,” we can see the degrees of “up” and down”—here over a range of 75 percent between 1995 and 2008, and everything in between. [visual]

So the question becomes: How valid is a constant number as a proxy for such an unpredictable range of returns in the wisdom of hindsight?

Let’s focus for a moment on VUL. How many of you were illustrating and selling these products, especially in the 1990s? And what illustration rate were you using? 10 percent? 12 percent?

Here’s an example I’ve charted for a VUL sold in the mid-1990s on a 43-year-old, healthy female. [visual] The solved for “What’s it gonna cost?” answer is a modest $4,062. Great! Sold! But let’s move ahead in time. The market performed remarkably until the beginning of 2000 when it hit some rocky shoals—what I call the “mama bear” market—and some extreme downward turbulence in the “papa bear” market of 2008‒2009. We don’t know how our hypothetical client would have reacted or allocated during these times, but let’s be really optimistic and assume she was able to average 10.5 percent and we’re reassessing in 2005. Just that 1.5 percent average difference between assumed 12 percent and assumed 10.5 percent return would have lapsed the policy by age 80. And if she hadn’t been paying attention—or didn’t know how to react to annual statements—what do you think the remediated planned premium calculation just prior to the lapse would be, with a more realistic average projection of 8 percent, and sustaining to age 100? $60,000 a year!

Many producers who were licensed to sell VUL moved away from this product after that “mama bear” market, concerned that VUL was too risky. Looking for alternatives, and prior to the move to indexed products later in the decade, no-lapse UL became the product of choice. Then, in this quick review of the evolution of universal life through the timeline, the Great Recession hit, and there was new interest and excitement in policies that could illustrate market-like returns but still feature a minimum guarantee regardless of depressing news from Wall Street. With official sales records from the last reporting year of 2016, IUL commanded a substantial part of universal life’s 36 percent sales by premium, neck and neck with whole life. By contrast, VUL was just 6 percent of the market by premium. So let’s consider: What crediting rate should be used to calculate planned premiums for IUL?

First, it’s important to look at some history of the last 40 years of the market as measured by the S&P 500. What do you notice about the ups and downs? [visual] Far more ups than downs, and that’s been fairly typical of a long view of market activity. So what does IUL do for the measuring stick of applicable crediting rates for IUL? It puts in a floor and a ceiling, or cap, possibly further constrained by a participation limit.

Technically, this is referred to as a “collar”—that within the hedging strategies pursued by the insurance company, still investing conservatively in bonds and other fixed-return assets, but taking excess premium not needed for the 0 percent guarantee to buy options in order to credit between 0 percent and the current cap/participation rate on a yearly calculated (typically) 365-day, point-to-point time frame.

So, to put this in context and perhaps probe a little on the client’s suitability for IUL versus VUL, we’re potentially going to give up all of this upside that might have been available in an S&P 500 index subaccount in exchange for being protected against the downside.

Let’s consider some details about point-to-point variability. Do you know how the S&P 500—the most-used index for IUL according to S&P—performed, for example, in 2010? With dividends, it was up 15.1 percent for the calendar year. Since the dividends that are part of the calculation of returns on this index are excluded, the 15.1 percent became 13.27 percent. And if the cap for that year had been 11 or 12 percent, then this distinction of stock dividends wouldn’t have mattered—the crediting rate would have been capped. But let’s keep digging deeper by focusing on August 2010 and the 365-day look back into August 2009. In the span of a little more than three weeks—depending on the anniversary date of the policy—the point-to-point differed by a multiple of sixfold. We read in the paper on New Year’s Day that the S&P 500 was up, in 2010, 15.1 percent. Consider the disparity and confusion for the policy owner whose measuring date was August 26. I say this not as a takeaway from IUL but to demonstrate the complexity of the product.

Let’s move into the policy illustration issues for universal life in general, and index policies in particular. I’m sure every one of you in your career has been the victim of the illustration beauty contest. You know, that’s where you propose a policy with a calculated planned premium only to be outgunned by another agent’s proposal for the same death benefit but a lower premium. The prospect thinks she has wisely found a better deal, not realizing it’s the same policy for which the only differences are the illustration projection assumptions. And, of course, you didn’t get the sale. Here are the illustration factors that won the beauty contest—and, most often, we’re hard-pressed to explain how the policy works and why $8,797 isn’t necessarily better than $10,500 and rightfully recapture the sale. [visual] So let’s apply some technology that’s readily available to financial planners, calculating the probability of success of meeting retirement income objectives, and run some new, hypothetical illustrations with randomly selected historical returns as an alternative to the AG 49 limit, in this case to the 6.48 percent constant crediting rate assumption. Here we’re going to run 1,000 hypothetical illustrations, simply testing the sufficiency of that proposed $8,797 planned premium.

Let me show you how we do this. We’re going to look at a graphical representation of the illustration proposing a planned premium of $8,797—first, confirming we’re incorporating a constant 6.48 percent crediting rate assumption and, then, showing the resulting development of cash value (in brown) and diminishing net amount at risk (in green) to the point where at age 100, the cash value equals the death benefit. [visual] These are the underlying assumptions of the sales illustration. Next, we’ll push the “random” button and generate for each of the next 53 years a random crediting rate derived from a database of almost 16,000 historic 365-day point-to-point returns since 1950.

What do you notice in this chart? One thing I see is that we predominantly have “full” cylinders—picking up the entire year’s return to the cap—or “empty” cylinders, representing fewer years of negative returns, which are calculated as 0 percent in the illustration. [visual] Now let’s see how the hypothetical cash value and net amount at risk accumulate under this first set of random assumptions. [visual] Not too bad, but it doesn’t make it quite to age 100, and it doesn’t diminish the net amount at risk the way our illustrated expectation suggested a few slides back. In fact, let’s bring that illustrated expectation back into the picture [visual], and we see that the randomly generated portrayal actually did better for a number of years but then hit a number of less-than-ideal returns in later years as the cost of the net amount of insurance began to increase in earnest, as the probability of death for each year increases substantially.

Another push on the “random” button gets us a less favorable view, but it’s just 1 out of a 1,000 of random portrayals of possibilities. [visual] So what do you think? When we look at a summary of 1,000 such randomizations, a meaningful statistical sampling, astonishingly calculated in under eight seconds, what’s the probability of success for a planned premium to sustain a policy to age 100 based on a conservative-sounding crediting rate as an average return for the next 53 years for this 47-year-old healthy male?

Well, the good news is that half the hypothetical, randomly generated illustrations made it to age 100 with “endowed” cash value. [visual] But the bad news is the other half did not make it. Confirming that we’re testing the proposed planned premium of $8,797 calculated at 6.48 percent, and a current cap of 11 percent with 100 percent participation, and accepting the calculated discount of lifetime illustrated expenses of 19 percent below industry average, we can now begin to assess the answer to “What’s the right premium?” in a brand-new way. This is not on the basis of an assumed crediting rate but on the buyer’s consideration of his or her risk tolerance translated to “What’s your minimum probability of success for this life insurance policy?” [visual]

This is where we totally change things around and have the client involved in setting the conditions for success. [visual] If she said 80 percent, we can calculate for you the likely initial planned premium that can fulfill that expectation. [visual] 90 percent? Sure, we can calculate that. And 100 percent? Well, I actually calculate to 99.9 percent just to discourage a consideration this represents a guarantee. It doesn’t. In fact, it’s not even accurate! Because we don’t know how future market conditions will affect this policy, we consider these calculations simply part of setting initial expectations. We’ll reapply the calculations in another three or four years, and with actual experience in accruing cash value in reference to an uncertain stock market, we’ll know whether with the higher planned premium feeding policy cash value has performed above the curve, on the curve, or below the curve and make appropriate midcourse corrections.

Notice that I stipulated and accommodated some assumptions from the original illustration—our calculation of expense deviation, the stated cap rate (which, by the way, is guaranteed at 3.5 percent and is changeable by the carrier as market conditions might change), and to endow at age 100.

Well, it appears that we’ve dropped to a one-in-three likelihood of success. [visual] If our client didn’t like 50 percent, she’s not going to like 32 percent. But we can go through the risk tolerance categories again, observing the progression of calculated initial planned premium for an 80 percent success requirement, or 90 percent, or 99.9 percent. At least, so far, these calculated initial planned premiums aren’t significantly higher than with the 11 percent cap assumption, but clients are starting to get a feel for how important their considerations are for personal risk tolerance—about something as important as a family’s financial sufficiency in the event of premature death.

Our final “test” will be to assume away this policy’s 19 percent lifetime lower expense assumption from that of the industry average and once again measure the calculated probability of success. This is not going to be acceptable, and yet there was no way to infer it from the underlying regulated illustration. We can progress through the probability thresholds and have a meaningful discussion with the client about how to best initiate and manage this policy to its successful conclusion. And it’s the client’s choice. The client could even decide to go for the aggressively low premium but would have made an informed decision that does not reflect badly on the agent!

We’ve seen the problem of projecting a constant crediting rate assumption against a policy style that depends on volatile results—even just within the collar of 0 percent and a current 11 percent—for a policy designed for lifetime protection. But the proposals we’re seeing more and more are, in fact, designed for a different purpose—where death benefit is just the feature enabling some significant tax benefits around accumulating cash value and later distributing it like a private defined pension plan. So let’s consider how our process might evaluate such a proposal. We’re going to take the exact same policy we’ve been reviewing, but illustrated as a private pension. We’ll use the illustration to set up cash flows of $72,500 a year for 15 years between ages 67 and 81, using that seemingly conservative crediting rate of 6.48 percent, all for a planned premium of $50,000 paid for each of the first 10 years. That’s $500,000 in for a designed $1,087,500 back out—tax free!

Here’s our chart that simply reflects the data points of the resulting illustration—premium paid for 10 years, loans and withdrawals, and tracking net death benefit. [visual] Here’s a view of the first randomization. By the way, if this were the actual result, what’s the tax impact of $500,000 in/$1,087,500 out, and a lapse right after the 15th cash flow distribution?

Here’s the tabulation of 1,000 randomizations similar to what we looked at earlier for the protection policy. [visual] We’re testing the calculated planned premium of $50,000 a year for 10 years and the 11 percent cap rate. Assuming this now familiar feedback of a 50 percent chance of success, what could we do to head off disappointment—if not lawsuits? Assuming we accept the premise of a constant 11 percent cap, randomization of returns between the collar of 0 percent and 11 percent with a 90 percent threshold of success, this suggests we need to plan on putting in $62,500, and not $50,000, and test again in another few years as we accumulate some actual cash value.

Once again, if we lower the cap expectation to a constant 10 percent, the result is likely unacceptable, and if this is important to the client, we might suggest she consider starting with an annual premium of $67,500.

Some IUL policies promote a non-guaranteed cap even higher than 11 percent. We’ve certainly seen some with 12 percent or more, and, of course, we can run a similar analysis. With a 12 percent cap, the planned 10-year premium could be as low as $37,860, but, as you are now conditioned to expect, there is not a great likelihood of success. What’s an appropriate adjustment? Back up to $50,000, which could have been approximated with a lower assumed crediting rate in the illustration—here of roughly 70 basis points lower at 5.8 percent. [visual] And while we won’t take the time here, obviously we’d want to test lower caps to give the client a more practical view of illustrated expectations versus probable reality.

Some critics of IUL have used our examples to suggest that there are too many moving parts to IUL and that it’s a much more risky policy design than most clients could match to their own risk tolerance.

But my own view is that we shouldn’t reject IUL because of its complications. I think we need to do a more detailed review of how these policies work so that we can test their suitability to our specific client and then offer credible, initial estimates of premium recommendations that comport with the client’s risk tolerance.

Let’s shift away from the numbers to acknowledge there’s a wide variety of policy styles that can fulfill virtually anyone’s needs and resources to pay for life insurance. If it’s for short-term protection—or if the budget for needed protection is strained—then term insurance is going to be the answer. As we get into the various needs and resources to pay for lifetime coverage—protection, cash flow in retirement, or estate liquidity, to name just a few—it comes down to how would we reasonably suggest UL or VUL or IUL or no-lapse, or whole life to a client? And is one style of life insurance better than another?

Here you’ll see a process we created to objectify recommending one type of lifetime “use” policy over another. [visual] If the client has a self-declared conservative risk tolerance for investing money—likely stated as “I am intolerant of volatility and seek guarantees”—then he or she is likely to feel best about lifetime use policies incorporating a high degree of guarantees, in other words, whole life and guaranteed death benefit (no-lapse) universal life.

A balanced risk style is likely to point to first-generation universal life, where the policy has a minimum guaranteed crediting rate with the possibility of higher credits based upon returns in the carrier’s general account assets. Such clients might say that they were “tolerant of modest volatility and willing to accept fewer guarantees in favor of premium payment flexibility.” Here the emphasis is on flexibility of premium payments.

A more aggressive risk tolerance, on the other hand, may point to a policy such as variable universal, in which the risk statement is likely to be: “I am tolerant of volatility and willing to do without guarantees in favor of premium investment opportunity.” Here, it’s important to point out that credits are based on chosen subaccounts. And, by the way, it’s important to point out that variable policies are not intrinsically risky! The policy owner has a wide range of subaccount styles from which to choose—from guaranteed fixed accounts to accounts that are designed to fit the requirements of conservative, balanced, aggressive, and very aggressive asset allocations.

We go on to indicate with all UL policies that premium sufficiency risk is transferred to the policy owner. Further, it’s important to point out that illustrated premium credits are assumed averages, usually unrelated to the ultimate selection of subaccounts!

Well, I’ve run out of traditional labels for risk tolerance, but there’s still one very popular policy style we haven’t gotten to. For this, my psychologist wife and I came up with a risk label to describe the risk statement of “I’m intolerant of volatility but drawn to the idea of upside with no downside.” We call that passive-aggressive. And, once again, we do a suitability review to make certain the client’s time horizon, investment savvy, resources, risk tolerance, tax bracket, etc. fit the profile for this type of policy.

This next slide is a one-pager I use with clients or producers we’re helping to bring home the point that we should carefully choose the illustration crediting rate by some process that fits the client’s needs—not just the highest rate the carrier will allow in the illustration software that would, in turn, suggest the lowest possible planned premium. [visual] Hopefully, I’ve effectively demonstrated the supplemental information not typically conveyed in an illustration—that high assumed rates, translating to low planned premiums, are a recipe for disappointment and subsequent customer complaint or lawsuit.

To complete this discussion of incorporating a new way to consider policy and illustration suitability, I want to acknowledge the reality that universal life has set up for us a process best described by Walmart’s founder Sam Walton: “Always low prices!” The expectation that the best policy is the cheapest policy is understandable when you consider that we tend to reduce buying decisions to price, especially when the technical aspects, whether of big-screen TVs or life insurance policies, are complicated and not readily appreciated other than by appearance.

So, here’s my recommendation: Run through the suitability process to pick an appropriate policy style for the client. Then, anticipate the illustration beauty contest and run two sets of illustrations on the chosen product: one at the aggressively higher crediting rate allowed by the illustration software, and one with a crediting rate appropriate to a high probability of success.

After running those two illustrations to calculate the respective planned premiums at max rate and, let’s say, 5.3 percent, I’d write in big, red letters on the top page of the illustration with a planned premium solved with the high crediting rate: “This illustration says I could choose to pay as little as $8,797 a year for this policy, but it’s not only NOT guaranteed; there’s a low statistical probability that the policy will ‘live’ as long as I do!”

On the second policy illustration, with the lower crediting rate, which in turn solves for a high initial planned premium, I’d write, “This illustration confirms my decision to pay $10,500 a year for my policy, because I’ve learned that I have a high probability the policy will outlast me. If future economic conditions are better, I can pull back my premium later on. If I have to put in more, it’s likely to be modestly more rather than disastrously more.”

Then, put both of those illustrations in an envelope or a file folder labeled “In case of buyer’s remorse—Open this file!” Because, you see, as good a job as you may do to contextualize why lowest premium is not the best way to pick a policy, the client will, in all likelihood, forget all that good logic after a while and possibly be vulnerable to a subsequent sales pitch by an aggressive agent: “If I could show you a way to have the same amount of life insurance for the same outlay, or more insurance for your current outlay, is there any reason we couldn’t do business together?” When the client, who has succumbed to the price logic of the pitch, goes to look for information about his or her policy, hopefully finding that file or envelope and the technique I’ve described, he or she will slow things down until you can sit down to validate why the client made the original decision.

Let me also say that you can’t buy the calculator that produces the probability figures. But it is an included member benefit of the Society of Financial Service Professionals, and membership eligibility has recently been expanded to include insurance and investment producers who do not have designations. You can get more information about this at

Richard M. Weber, MBA, CLU, is a 50-year veteran of the life insurance industry, having been a successful agent, an insurance company executive and now a consultant to insurers and their agents on the topic of effective selling. Weber is the author of more than 300 articles encompassing products, sales practices, and the due diligence necessary to buy and sell insurance and annuities. His research paper, “Life Insurance as an Asset Class,” co-authored with Chris Hause, received the 2008 Academy of Financial Services "Best Paper” award.


{{GetTotalComments()}} Comments

Please Login or Become A Member to add comments