Ben Mathew wrote: ↑
Sun Aug 02, 2020 12:05 am
petulant wrote: ↑
Sat Aug 01, 2020 6:31 pm
Ben Mathew wrote: ↑
Sat Aug 01, 2020 4:37 pm
Rick Ferri wrote: ↑
Fri Jul 31, 2020 7:48 pm
If you REALLY listen to what Dr. Pfau said, it’s to be agnostic. Don’t have biases. Be open to new ideas.
In the interview (about 45:00 min in), Pfau said that there are lower cost products available to fee only advisors. I'm intrigued. Given that you are a fee only advisor, it would be interesting to hear your thoughts on that.
I have been interested in this question too. I found an article the other day by Kitces discussing TIAA-CREF's no-load life insurance, but it appears they exited the business in 2019. Today I found a link to a company called Ameritas that has a "no-load VUL." It appears the only charges are a 2% premium tax, an administrative fee of about $15 per month, the cost of insurance, an annual face amount charge that is currently zero, and a mortality/expense charge starting at 0.70% per year for the first 15 years and then going to 0.10%. In addition to a wide variety of mutual fund options including Vanguard and DFA, it has a fixed account option guaranteed to earn at least 3%. Since the cash value can be removed with no taxes through a withdrawal of premium or loan, the 3% is a direct aftertax premium but after the .70% charge would be 2.3%. I am looking into it more just to understand it.
Thanks for the lead.
If somebody who is familiar with this type of insurance product can compare holding VTSAX directly vs through the Ameritas Advisor II Variable Universal Life
policy, please do.
I've modeled VUL policies before to figure it out, and it's hard for them to win. Looking at the VUL as a way to hold assets, the vehicle is driven by up-front and ongoing costs compared to tax benefits. For example, holding bonds inside the policy avoids ordinary income tax on the bonds, allowing them to effectively earn a higher return; it also allows all premiums paid to be included in basis, even though some of the premiums were used to cover cost of insurance. The vehicle also has harder-to-model tax benefits like withdrawing through return of premium or policy loan without generating income tax as well as being able to rebalance inside the policy without incurring taxes. Another aspect of the comparison is that VUL policies often include a "fixed account option" or guaranteed option akin to a stable value fund inside a 401(k). These currently have higher returns than bonds--the Ameritas VUL has a minimum 3% fixed option. But from a theoretical perspective, it is much harder for normal stocks to win inside a VUL since it trades QD tax rates and ultimate LTCG on liquidation for possible ordinary income tax on surrender.
This morning I finished a crude model to look at how the Ameritas product performs compared to a taxable brokerage account. I looked at a 35-year-old male who already needs $500,000 life insurance for 30 years and compared buy-term-and-invest-the-rest to VUL. In this example, the 35-year-old male invested $5,000 annually in investments+life insurance until age 65 and then surrendered/liquidated all amounts at age 80. I then studied how outcomes vary based on asset allocation and state (for calculating income tax differences). What I found was that the brokerage account always wins for a 100% stock portfolio, while the VUL always wins for a bond portfolio based on existing interest rates. Whether the brokerage account or VUL option is better for 50-50 depended on the tax rates applicable, with no-income-tax state investors in a low bracket actually coming out ahead with all 50/50 in VUL. I also assumed the investor only pays cost of insurance based on amount at risk with a fixed death benefit (unless increasing is necessary due to corridor using very crude assumptions).
This isn't perfect modeling because there are aspects that are very hard to model well. For example, it may be unrealistic to liquidate the entire policy at age 80; on the other hand, if COI continues to escalate, that might be the best option. But in the event of death, there is no tax. But there's not tax on the stocks at death either. Also, much of the advantage for VUL in lower tax brackets disappeared if I changed the bond return to equal 3%, though at higher tax brackets VUL is a clear winner for bond assets.
Further, many of the numbers were actually close enough that execution mattered more than the asset location. For example, in one scenario I looked at an investor from a state with an average income tax of 4% using a 50/50 asset allocation. By age 70, this investor would have $382K in the VUL with $155K in basis and thus a net value of $323K if the entire VUL was surrendered. By comparison, using a traditional brokerage account, the investor would have a $348K balance, $221K in basis, and $324K in value if all positions were liquidated. Moving past age 70, the tax deferral of the VUL will allow it to grow faster than the taxable brokerage account until COI becomes too large a drag in later years. The close performance near age 70 indicates that execution would probably make a big difference, however: during their 60s, the investor may be able to realize capital gains in the brokerage account more efficiently, increasing the basis; while the VUL investor would need to start winding down the policy to realize gains--taking a withdrawal up to the basis and then even more while reducing the death benefit to minimize COI based on amount at risk. Which way the investor is better off depends entirely on objectives and more particular circumstances.
Due to the complexity involved, I still think these should be reserved for high-tax-bracket investors who would otherwise hold bonds in a taxable account (after using up all tax-advantaged space), who have a need for life insurance anyway, who plan to exit the policy before aging too far along (since cost of insurance will eventually eat or drag on the policy), and especially who might have unique needs for asset protection or to avoid estate tax.
(Note, my model is not an actual illustration and should be taken with a grain of salt. I can share the model over PM by request.)