Flawed Study Captures the Variable Annuity Industry’s Attention

I wanted to provide some feedback on a recent working paper that was brought to my attention, before heading out for the weekend.

One of my favorite journalists contacted me on Wednesday, September 16, 2020- asking for commentary on “Conflicting Interest and the Effect of Fiduciary Duty – Evidence from Variable Annuities.” This is working paper 21-018, funded by Harvard Business School, and authored by Mark Egan (Harvard Business School), Shan Ge (New York University), and Johnny Tang (Harvard University).

You can download the working paper here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3652115.

As a preface, I do not endorse any company or product. I am channel-agnostic. I believe that any compensation method is valid and useful if it gets tools in front of consumers so that they can determine if the aforementioned tools help them by providing needed solutions. I run a third-party market research firm, Wink, Inc., which tracks life insurance and annuity companies, their products, features, rates, and sales. Honestly, I would never personally purchase a variable annuity; it does not match my risk tolerance. That aside, I have some thoughts on this working paper.

Note that I am no Harvard Business School graduate, but I am recognized as the nation’s top expert on annuity products.

There are some considerably amateur flaws with this “study.”

• It does not appear that differences in distribution channels were taken fully into consideration. While the authors repeatedly mention captive and independent salespeople, there is no mention of direct response sales, bank sales, nor Registered Investment Advisor (RIA) sales. How are sales of Variable Annuities (VAs) outside of these two distributions accounted for? Technically, every distribution short of captives can sell from multiple insurance companies’ product shelves. That said, what drives sales (i.e. commissions, bonuses, fees, etc.) varies significantly from one distribution channel to the next (i.e. credited rate in banks, commissions in independent agent, etc.).
• It is imperative to understand that commissions on two like annuities, one sold via a captive salesperson, and the other through an independent advisor, will differ dramatically. The captive salesperson is an employee of the insurance company; s/he is likely receiving benefits such as health/dental insurance, office space, staff, leads, advertising, and more. As such, captive commissions are typically a fraction of independent advisors’, who must provide for all of those perks on their own. In addition, products sold through independent advisors have an additional layer of compensation built into them, to compensate the distributor. So, relatively speaking, the authors are comparing apples to oranges; the commissions cited in a prospectus for a captive agent are compensating a single person where the independent agent prospectuses account for compensation for two parties. The authors seem not to grasp these basic, fundamental differences in annuity pricing.
• The authors compare sales from 2016 to 2017 in their study. The Department of Labor’s “fiduciary rule” was proposed on April 6, 2016. This means that there was a full quarter of “business as usual” annuity sales taken into consideration for the year 2016, and unaffected by a massive proposal that would administratively alter sales in a significant manner.
• Sales of ALL annuities declined in response to the Department of Labor’s (DOL’s) fiduciary rule (declines ranged from – 7% drops to -29% drops, based on the type of annuity; fixed, variable, indexed, etc.). This is because the burdens of the rule required technological changes, operational changes, product management changes, and sales process changes (just to name a few points-of-impact). These massive changes distracted ALL insurance companies, selling ANY type of annuities. So, sales declined because insurance companies were distracted from the business of marketing their products, and instead focused on executing major overhauls to their businesses.
• LIMRA’s sales data indicates that VA sales declined 9% from 2016 to 2017; not 19% (https://www.limra.com/globalassets/limra/newsroom/fact-tank/sales-data/2017/q4/4q-2017_annuity-estimates.pdf).
• The authors indicate that they identified variable annuity commissions based on information in prospectuses. The commissions that are filed in prospectuses are extreme examples of the top range of commissions. Filing a prospectus is time-consuming and expensive. Insurance companies want to maximize the flexibility of their product filings, in preparation for potential future changes to the market environment, whether positive or negative. As a result, the commission amounts cited in prospectuses are rarely the amounts actually paid to salespeople. This working paper cites a range in available VA commissions from 0% – 16%, which is misleading. I am aware of VAs paying as high as 8% commission, not any greater, and certainly not 16%.
• It is also relevant than the commission amount that is paid to the independent advisor is not determined by the insurance company; the Broker/Dealer (B/D) holds that authority. This is because the insurance company pays the B/D, who takes their override commission, and then passes on what they deem to be sufficient to the independent advisor. So, the commission amounts cited in this paper are what is called Gross Dealer Concessions (or the all-in compensation); it is not the amount paid to the advisor.
• Further, the authors appear ignorant of the fact that many B/Ds moved to a levelized commission, across all share classes, because of the DOL’s proposed rule. As such, the same annuities likely paid greater commissions in 2016, as compared to the following year.
• The authors seem to be under the impression that the insurance company determines the expense ratios on VAs in a vacuum, when in fact those fees are determined by the fund manager in conjunction with the insurance company.
• The authors make an egregious mistake in comparing the 2% median commission on mutual funds to the nearly 7% average commission on VAs. There is a slew of no-load mutual funds, which skew that statistic. On the other hand, only 8.20% of VA sales are fee-based/no load.
• The authors seem to lack a basic understanding of the differences in variable annuity share classes. A comparison of C- and L-share products, offered by [an unnamed large insurance company that is a major contender in the annuity market], is illustrated; inferring that the C-share product is in the advisors’ best interests, rather than the purchasers’. However, the authors appear ignorant of the fact that C-shares lack surrender charges and are therefore desirable for those seeking potential liquidity. As a result, C-shares typically have slightly higher fees, relative to other share classes. On the other hand, L-shares DO have surrender charges, and subsequently (relatively) lower fees. The inference that the L-share/surrender charge sales being significantly greater than C-share/liquid annuity sales, simply because it has higher fees (and therefore higher commissions) is erroneous. It is also worth noting that the reduction in sales for “high fee” VAs was more a result of a great many L-shares being pulled from B/Ds shelves because of suitability (i.e. lifetime riders with short surrender schedules); this was the B/D’s choice to remove such products from the platform, and not something controlled by the advisors.
• As a result of the authors’ ignorance of share classes, assessments of differences in share classes are imminent. For example, X-shares have higher/longer surrender charges than other share classes by virtue of offering a bonus to the purchaser. Incidentally, every X-share VA available today is sold by insurance companies that distribute through independent advisors. Is this taken into consideration in the analysis of fees? It is hard to conceive that such differences could be accounted for.

There are a great number of conclusions drawn which could possibly be correlated, but not necessarily causal, in this working paper.

• Captive salespeople do not necessarily place a “higher weight on their clients’ incentives” relative to independent advisors. Generally, captive salespeople have access to sell just a few, very different, variable annuity products (i.e. a C share, an L share, and an X share). Alternatively, independent advisors have access to sell 211 different variable annuities, each with different features. Independent advisors may have access to 6 different A-shares, 122 different B-shares, 29 different C-shares, 54 different I-shares, 26 different L-shares, 13 different O-shares, and 10 different X-shares. This can result in a wide range of variability in product features, including expenses.
• Complaints should be greater for products with higher fees, as generally the products with higher fees are sold by independent salespeople. Products available through independent advisors generally have higher commissions and higher fees as a result (see previous note on differences in distribution compensation), as compared to captive salespeople and the products available to them. By virtue of distribution, captive salespeople have better training, handholding, and relationship-orientation than their counterparts. Independent advisors have comparatively less oversight, little assistance, and more transaction-orientation than the captives (relatively speaking). So, the issue at hand is a matter of distribution, not necessarily product.
• It should also be taken into consideration that B/Ds began scrutinizing the suitability of exchanges more closely, post-DOL. This too would result in a lower level of complaints after the rule was proposed (as well as lower sales).
• If anything, the findings in this study seems to support an observation “captive salespeople experienced lesser disruption of their VA sales than their independent counterparts, as a result of their captive home office dictating what changes to make as a result of the DOL’s rule; whereas independent advisors were often left questioning what changes were required to comply with the fiduciary rule and confused by conflicting guidance offered by multiple insurance home offices that they have relationships with.” It would be impossible to draw the conclusions that the authors did, reliably, with such flawed assumptions throughout the working paper.

Ultimately, after reviewing this working paper, I am left feeling as if I just finished reading 57 pages of propaganda; bought and paid for by Congressional Democrats. Do not get me wrong, I AM a conservative Democrat. However, I must call it like I see it.

I welcome the opportunity to assist anyone in the academic community, in their efforts to analyze the proposed fiduciary rule’s impact on the annuity industry. I believe that the mistakes noted above can be eliminated if working with a disinterested expert. As a result, my door is open. Just call me at (515) ANN-UITY.

All of that said, I would also like to disclose my personal opinions on the implementation of a “fiduciary rule.”

Fiduciaries have been trained for a VERY long time about how “terrible” annuities are for consumers (for a number which are outside the scope of this blog). So, the RIA distribution channel not only has to be educated about the realities of this product but also needs to be grown into something significant. In short, all of the fee-based annuities in the world will not help us, unless we have people to sell them. Further, I do not believe any annuity is “bad” OR “good”; products are merely tools. Now, is there such a thing as an unsuitable sale? Absolutely. However, that occurs when the salesperson does not appropriately identify the right tool to solve the problem.

Most Registered Investment Advisors (RIAs)/“fiduciaries” will not work with clients who have assets of less than [$250,000]. Sadly, the average annuity sale is less than $200,000. So, RIAs are not the ideal demographic for best serving the people who need the guaranteed income floor the most (a feature provided by annuities). The reason the RIA requires a minimum asset level of [$250,000] is so that s/he does not have to meet with hundreds of clients, to make ends meet. However, this phenomenon also results in middle America being left in the dust, should a fiduciary rule be enacted. The #1 fear of Americans is running-out-of-money in retirement. Annuities are strategically positioned to address this need. So, how would the right solution get into the consumers’ hands, without the aid of a financial professional that is comfortable with the aforementioned solution?

The lack of choices for fee-based annuities concerns me; only 83 of the nearly 1,500 annuities in Wink’s AnnuitySpecs tool are fee-based, rather than commissioned. And, there will not be more fee-based annuities unless there are more sales of fee-based annuities. Fee-based annuity sales only account for 8.68% of deferred annuity sales as of 2Q2020 (Wink’s Sales & Market Report). We will not have more sales of fee-based annuities until we find more RIAs who are open to selling annuities. Quite the dilemma…

While a fiduciary world may sound great to some, it alienates the 99% of annuity salespeople who are NOT RIAs. And by virtue, it also alienates the clients of said salespeople. THAT is the rub.

For this reason, I believe we need both commissioned and fee-based options in the annuity industry. Many other industries have multiple compensation options. For example, you can purchase a car through an auto dealer by working with one of their salespeople on a car lot; where a commission is paid. You can purchase a car through an auto dealer by working with one of their salespeople online, where a flat fee is paid. You can purchase a car through an app, and not deal with a salesperson or compensation at all. None of these methods are “bad,” they are just different. And you are never going to convince my grandma to buy an $80,000 vehicle on an app through her phone. Alternatively, a millennial would probably prefer not to invest the time into spending hours with a salesperson on a car lot, when they can get what they want instantly via their phone.

Unless someone can solve these problems in a workable fiduciary rule, I will not support the regulation.

This article originally appeared on Wink Intel.