Historical Whole Life Comparison

Over the course of the past several weeks, we looked at real whole life policies and evaluated their real policy performance.  We specifically looked at cash value accumulation comparing original projections of non-guaranteed values to the actual performance of the policies.  Every policy spans a timeline of around 10 years.  we used internal rate of return calculations to control for variations in premiums paid.

We were able to review five different whole life products from five different insurance companies.  Products and companies includes were:

All five whole life insurance policies used policy blending and paid-up additions to augment cash value performance.  In other words, they all made use of design tweaks that seek to increase the cash value performance of the whole life policy.  We put in force and manage four of these policies.  One of them (the Northwestern Policy) we did not put in force.

We made some interesting discoveries with this review.  Notable observations included things like: changes in the projected dividend do not necessarily correlate with changes in accumulated cash value, the company that performed closest to its projections was still not the best performer for overall rate of return achieved, and there is a very wide range in terms of results.

Here's a table that summarizes the results:

Company Projected Actual Dividend Change
Guardian 1.66% 1.63% -72%
MassMutual 4.65% 2.97% -42%
Ohio National 0.95% 0.15% -46%
Penn Mutual 5.39% 5.11% -24%
Northwestern Mutual -1.05% -4.45% -10%

Dividend Changes Didn't Predict Cash Value Changes

Whole life is a unique product from one company to another.  We've cautioned this point in the past.  The important take away to realize that you cannot assume that whole life insurance at one company is identical to whole life insurance at another company.  If you assume that all whole life insurance is universally the same, you could end up buying an insurance policy that is not at all equipped the meet the demands of your specifics goals.

The greatest way I can think to quantify this point is by looking at how these whole life policies performed while also looking at the change in their dividend.  The change in dividend is in the most recent policy year.  So, we calculated the change in dividend relative to the originally projected payable dividend in that policy year.  For example, if the original projection said the company would pay a $5,000 dividend in year 10, but the actual dividend payable in year 10 was $3,500, the reported value in the “Dividend Change” column is -30%.

Notice that the company with the closes actual result to the original projection is Guardian, but it also has the largest change in payable dividend.  Notice also that the company with the largest change in actual cash value to projected is Northwestern Mutual, but it has the smallest change in payable dividend.

Most of us–me included–would reasonably assume that changes in dividend would positively correlate with changes in cash value.  But that's not the case.  We can explain some of this with paid-up additions and the time elapsed since inception.  But the bigger explanatory variable is simply that all whole life is unique and the pieces and parts differ wildly in terms of how they function from one company to another.  There is no unification for the functionality of a dividend on a whole life insurance policy.  How insurers choose to treat them and make them function on a policy is largely up to their discretion.

Having Accurate Projections Doesn't Mean Better

While we can applaud Guardian for having the most accurate result compared to its original projection, there are still two whole life products that performed better.  Both MassMutual and Penn Mutual produced a higher rate of return on cash value than the Guardian policy.  MassMutual experienced considerable deviation from its original projection but still turns out as a better deal for the cash value-maximizing seeker.

Some people have considerable concern over the accuracy of a cash value ledger produced by a life insurer.  I understand the concern.  I've also pointed out on a few different occasions that when we have one result substantially higher than the other, we have to contemplate just how much the higher projected company has to be wrong to match the next option.

Using this logic and focusing on MassMutual and Guardian for a moment, we see that MassMutual needed to be wrong by almost 300 basis points to match Guardian.  The actual results from MassMutual are considerably below projection, but it only fell by slightly more than half that amount making MassMutual still the better deal rate of return wise.

Considerable Variation Among the Whole Life Insurance Products

All of these products are around the same age in terms of time in force.  We have a maximum result of 5.11% rate of return from Penn Mutual and a minimum of -4.45% rate of return from Northwestern Mutual.  For those looking at blog posts/articles written by others who seek to argue for or against whole life insurance, you may now understand why we dismiss their work when they use one example of a whole life ledger to support their position.

This speaks yet again to the fact that whole life varies from company to company and its uniqueness at each company makes it better or worse for specific applications.  The decision to buy whole life insurance is an individual one based on specific needs and it cannot be decided through arbitrary boundaries one uses in a hypothetical circumstance to blanketly declare it's either all good or all bad.

6 thoughts on “Historical Whole Life Comparison”

  1. So, based on your analysis are you saying the Penn Mutual returned the greatest dividend over the pas 10 years of all the company’s you measured above?

    • Hi Jerry,

      That’s not exactly what the article/podcast is saying at all, though it may be technically correct depending on how you characterize “greatest dividend”. The analysis is saying that all of the companies have underperformed their original illustrated return on cash over the last 10 years. Some have done way worse (Northwestern) while Penn Mutual has underperformed the original illustration but not by much and that specific fact can be attributed to their lower dividend now than it was 10 years ago.

  2. Can you clarify what are columns 2 and 3? IRR? If it is IRR, then the holder of northwestern policy has been losing more money than projected?

    • Hi Jay,

      Yes, columns 2 and 3 are showing the internal rate of return on cash value (IRR). And yes, the holder of the Northwestern policy has lost more money than was originally illustrated over the ten-year period.

  3. It appears that you do your homework but still find comparisons less than definitive when it comes to judging various companies. You do seem to prefer the mutual companies that pay dividends. Perhaps you should look at how dividends arise; a savings in mortality and expenses, investment results, and lapse ratios. Compare how the companies perform in these areas and put less emphasis on the short term dividend payments. You may want to look into complaint ratios which are published by many states. Complaint and lapse ratios would certainly provide some insight in to how policy owners feel. All companies collect premiums and investment income, and pay their expenses and death claims. What’s left is paid out as dividends to policyholders or stockholders. In other words, you’ve considered dividends paid.
    I believe it would be very helpful and informative to take a look at the underlying factors. You are up to it!

    • If the takeaway from this is we focused on dividends paid, you either didn’t read it or you misunderstood it. It’s about summarizing policy performance over the time period. What actually happened. We used dividends simply to point out that the concept differs wildly across companies in terms of what dividends do to policy performance. We took this on specifically because many insurance marketers–both agents and companies–have placed an unrelenting focus on dividends, which often means very little big picture.

      We didn’t intend for this to be a definitive study on who is the best. It’s merely a look at real policies and how they performed.


Leave a Comment